Scaling and intermittency in Burgers turbulence

J. P. Bouchaud, Marc Mézard, Giorgio ParisiPhysical 1995 · DOI · arXiv

Abstract

The problem of fully developed turbulence in the $N$-dimensional forced Burgers equation is investigated through a mapping to directed polymers in random media. By analyzing the polymer's behavior on length scales smaller than the correlation length of the random potential, we propose an ansatz for the velocity field in the large-Reynolds-number limit. This approach, which becomes exact as $N \to \infty$, allows for the derivation of the full probability distribution of velocity differences. We find that the scaling exponents $\zeta(q)$ collapse to unity for all $q \ge 1$, a result of the "froth-like" shock structures that dominate the field.

Introduction

Turbulence is often described as the last great unsolved problem of classical physics. While Kolmogorov’s 1941 (K41) theory provides a foundational framework for the energy cascade, it assumes a degree of spatial homogeneity that is contradicted by experimental evidence of intermittency. Intermittency refers to the presence of rare, intense fluctuations at small scales that deviate from Gaussian statistics.

To better understand these phenomena, researchers often turn to the Burgers equation. Lacking the pressure term and vorticity of the Navier-Stokes equations, the Burgers equation nonetheless retains the essential non-linear competition between convection and viscosity. In its forced form, where a random stirring force maintains a steady state of turbulence, it serves as a powerful laboratory for studying the geometric origins of intermittency—a regime often called "Burgulence."

Methodology: From Fluids to Polymers

The central innovation of this work is the application of a "dictionary" between fluid dynamics and statistical mechanics. The mapping is achieved through two primary steps:

  1. The Cole-Hopf Transformation: This standard mathematical tool maps the non-linear Burgers equation to a linear heat equation with a random potential. Physically, this corresponds to the partition function of a directed polymer—an elastic string—subjected to a quenched (spatiotemporal) random potential.
  2. Replica Symmetry Breaking (RSB): To solve the disordered problem of the polymer in a random medium, the authors employ the Replica Method, a technique originally developed for spin glasses. Specifically, they use a One-Step Replica Symmetry Breaking (1-RSB) variational ansatz. In the context of the fluid, the "replica" states represent the various metastable configurations or local energy minima that the system can occupy.

This 1-RSB ansatz is particularly powerful in the Large-N limit (where $N$ is the number of spatial dimensions). While the mapping is exact for $N \to \infty$, the physical insights remain relevant for lower-dimensional systems.

The Geometry of "Burgulence"

The analytical solution reveals a distinct geometric structure for the velocity field. Rather than the complex, multi-scale vortices seen in Navier-Stokes turbulence, the Burgers velocity field organizes into a "froth-like" structure.

Within the "cells" of this froth, the velocity varies smoothly and radially. However, these cells are separated by $(N-1)$-dimensional boundaries where the velocity jumps abruptly. These boundaries are shocks—singularities where the kinetic energy of the fluid is dissipated by viscosity. In this model, intermittency is not a subtle correction to scaling but a dominant feature driven entirely by these shock structures.

Scaling Laws and Velocity Statistics

The 1-RSB approach allows for the exact calculation of the probability density function (PDF) for longitudinal velocity differences $u(r)$ at distances $r$ much smaller than the forcing correlation length.

The Collapse of Scaling Exponents

In standard K41 theory, the $q$-th order velocity moments are expected to scale as $\langle u^q(r) \rangle \sim r^{\zeta(q)}$, with $\zeta(q) = q/3$. In Burgers turbulence, the authors find a dramatic "collapse": $$\zeta(q) = 1 \text{ for all } q \ge 1$$ This implies that the statistics are dominated by the discontinuities (shocks) rather than the smooth flow within cells. Interestingly, at $q=3$, the theory yields $\zeta(3)=1$, which is consistent with Kolmogorov’s "4/5 law," suggesting that the energy cascade mechanism in Burgers' equation shares fundamental similarities with real turbulence despite the difference in intermittency.

Energy Spectrum and Tracers

The energy spectrum is found to follow $E(k) \propto k^{-2}$ for all $N$, a signature of a field dominated by shocks. Furthermore, the study examines the behavior of passive tracers (e.g., dust particles) within the flow. Due to the dissipative nature of the shocks, tracers tend to aggregate into high-density "sheets" at the shock locations, leading to anomalous fluctuations and power-law tails in concentration distributions.

Conclusion and Limitations

While the Burgers equation provides a tractable model for intermittency, it differs from Navier-Stokes turbulence in key ways. The most significant is the lack of vorticity; in real fluids, singularities are typically 1D vortex filaments, whereas in Burgulence, they are $(N-1)$-dimensional sheets. This explains the "stronger" intermittency observed in this model.

Future work may involve extending the 1-RSB ansatz to account for rotational flows or exploring "Full RSB" for lower-dimensional cases. Nonetheless, the mapping between directed polymers and turbulence provides a robust theoretical bridge, allowing tools from the physics of disordered systems to shed light on the complexities of fluid motion.

Summary updated

Abstract

We use the mapping between the Burgers equation and the problem of a directed polymer in a random medium in order to study the fully developed turbulence in the N-dimensional forced Burgers equation. The stirring force corresponds to a quenched (spatiotemporal) random potential for the polymer. The properties of the inertial regime are deduced from a study of the directed polymer on length scales smaller than the correlation length of the potential, which is not the regime usually considered in the case of polymers. From this study we propose an ansatz for the velocity field in the large-Reynolds-number limit of the forced Burgers equation in N dimensions, which should become exact in the limit N\ensuremath{\rightarrow}\ensuremath{\infty}. This ansatz allows us to compute exactly the full probability distribution of the velocity difference u(r) between points separated by a distance r much smaller than the correlation length of the forcing. We find that the moments 〈${\mathit{u}}^{\mathit{q}}$(r)〉 scale as ${\mathit{r}}^{\mathrm{\ensuremath{\zeta}}(\mathit{q})}$ with \ensuremath{\zeta}(q)\ensuremath{\equiv}1 for all q\ensuremath{\ge}1 [in particular, the q=3 moment agrees with Kolmogorov's scaling \ensuremath{\zeta}(3)=1]. This strong ``intermittency'' is related to the large-scale singularities of the velocity field, which is concentrated on an (N-1)-dimensional frothlike structure, which is in turn related to the one-step replica-symmetry-broken nature of the associated disordered problem. We also discuss the similarities and differences between Burgers turbulence and hydrodynamical turbulence and we comment on the anomalous tracer fluctuations in a Burgers turbulent field. Since this replica approach is rather unusual in turbulence problems, we provide all the necessary details of the method.
BibTeX
@inproceedings{bouchaud1995scaling,
  author = {J. P. Bouchaud and Marc Mézard and Giorgio Parisi},
  title = {{Scaling and intermittency in Burgers turbulence}},
  booktitle = {Physical},
  year = {1995},
  doi = {10.1103/physreve.52.3656},
}
Figure