Generalization of the viscous stress tensor to the case of non-small gradients of hydrodynamic velocity: a path to numerical modeling of turbulence non-locality

This paper generalizes the Chapman-Enskog method to derive an integral representation of the viscous stress tensor for large velocity gradients, enabling the numerical modeling of turbulence non-locality and phenomena like tangential discontinuities that standard Navier-Stokes formulations struggle to capture.

The Gist

The Big Problem: The "Smoothie" vs. The "Storm"

Imagine you are trying to predict how a fluid (like water or air) moves. For over a century, scientists have used a famous set of rules called the Navier-Stokes equations. These rules rely on a specific ingredient called the viscous stress tensor.

Think of this tensor as a "friction calculator." In the standard version, it assumes that if you push a fluid, the resistance (friction) it feels depends only on how fast the fluid is moving right next to the point you are looking at. It's like assuming that if you stir a cup of coffee, the resistance you feel is determined only by the coffee molecules touching your spoon.

The Flaw:
The author, A.B. Kukushkin, points out that this standard "friction calculator" breaks down when things get chaotic, like in a turbulent storm or when two streams of fluid slide past each other at high speeds (tangential discontinuities).

• The Analogy: Imagine a crowd of people walking in a hallway. The standard model assumes everyone only bumps into the person immediately next to them. But in a real crowd (turbulence), a person might get pushed by someone three rows back, or a wave of movement might travel across the whole room. The standard model ignores these "long-distance" interactions.
• The Paradox: The standard math also leads to a weird result: it suggests that if particles collide more often (like in a thick fog), the fluid should actually flow easier (lower viscosity). This feels backwards to our intuition.

The Solution: Looking at the Whole Picture

Kukushkin proposes a new way to calculate this friction. Instead of looking only at the immediate neighborhood, his new formula looks at the entire history and location of the fluid's movement.

The New Approach:

1. Abandoning the "Small Steps" Rule: The old math (Chapman-Enskog method) only works if the fluid changes very slowly and smoothly. Kukushkin removes this rule. He allows for sudden, sharp changes in speed, which happen in real turbulence.
2. The "Messenger" Analogy: Instead of just looking at the fluid at one spot, imagine the fluid is full of tiny "messengers" (vortices or disturbances).
   • In the old model, a messenger only talks to its neighbor.
   • In Kukushkin's new model, a messenger is born at one spot, flies across the room, and delivers its message to a spot far away before it stops.
3. The Integral Formula: The new math is an integral (a sum over a large area). It calculates the stress (friction) at a specific point by adding up the effects of all these messengers traveling from everywhere else in the fluid to that point.

Why This Matters

1. Fixing the "Paradox":
By allowing these messengers to travel long distances, the new formula fixes the weird paradox about viscosity. It explains why fluids behave the way they do even when particles are colliding frequently. The "long flights" of the messengers account for the resistance in a way the old "short steps" model couldn't.

2. Connecting to Real-World Chaos (Richardson's Law):
The paper mentions a famous observation called Richardson's $t^3$ law.

• The Analogy: If you drop two leaves into a turbulent river, the standard model predicts they will drift apart slowly (like $t^2$). But in reality, they fly apart much faster (like $t^3$).
• The Connection: This new "long-distance" model naturally explains why particles separate so quickly. The messengers travel far and fast, carrying the disturbance across the fluid, which matches the real-world observation of how turbulence spreads.

3. A Bridge to Better Computer Simulations:
Currently, computer simulations of turbulence often have to use "cheats" or made-up numbers because the standard math fails at sharp edges (like where a wing separates from the air).

• Kukushkin's new formula provides a mathematical bridge. It turns the "cheats" into a rigorous calculation based on first principles. It allows computers to model turbulence by summing up these long-distance interactions, rather than just guessing.

Summary in a Nutshell

The paper argues that the old way of calculating fluid friction is like trying to understand a conversation by only listening to the person standing next to you. It misses the big picture.

Kukushkin has written a new rulebook that listens to the entire room. By accounting for how disturbances travel across the whole fluid (non-locality), this new math:

• Fixes a logical paradox about how thick or thin a fluid should be.
• Explains why particles in a storm fly apart so quickly.
• Offers a path for computers to simulate complex, chaotic flows (like wind around a plane or water in a pipe) much more accurately, without needing to rely on guesswork.

The author notes that this same logic could eventually be applied to heat transfer and even plasma physics, but the core achievement here is rewriting the rules of fluid friction to handle the "messy" reality of turbulence.

Technical Summary

Technical Summary: Generalization of the Viscous Stress Tensor for Non-Small Gradients

Problem Statement
The paper addresses fundamental limitations in the standard Navier-Stokes description of hydrodynamic turbulence, specifically regarding the viscous stress tensor ($\sigma_{ik}$). The conventional tensor, derived via the Chapman-Enskog method from the Boltzmann equation, relies on the assumption of small gradients of hydrodynamic velocity. The author argues that this assumption renders the standard tensor inadequate for describing complex flow phenomena such as tangential discontinuities, separated flows, and the non-locality inherent in turbulence.

Furthermore, the paper highlights a theoretical paradox in the standard derivation: the dynamic viscosity coefficient ($\eta$) is found to be inversely proportional to the particle collision cross-section ($\sigma_{coll}$), contrary to intuitive expectations. The author attributes this paradox to the restrictive "small gradient" condition of the Chapman-Enskog method, which becomes increasingly feasible only as the collision cross-section increases, thereby masking the behavior of long-running localized vortices.

Existing turbulence models (e.g., Spalart-Allmaras, RANS) often resort to phenomenological corrections or local differential equations that fail to capture the non-local nature of turbulence, evidenced empirically by Richardson's $t^3$ law for pair correlations in turbulent media. While non-local transport models (such as Lévy flights/walks) have been proposed, they often lack a rigorous derivation from first principles within hydrogasdynamics.

Methodology
The author proposes a generalization of the viscous stress tensor by relaxing the condition of small velocity gradients within the Chapman-Enskog framework. The methodology proceeds as follows:

1. Kinetic Foundation: The derivation begins with the Boltzmann kinetic equation for a single-sort gas. The distribution function $f(\mathbf{p}, \mathbf{r}, t)$ is decomposed into a local thermodynamic equilibrium part (Maxwellian distribution $f^{(0)}$) and a correction term $\varphi$.
2. Relaxation of Gradient Constraints: Unlike the standard approach where spatial derivatives of the correction term are neglected, the author retains the derivatives of $\varphi$ on the left-hand side of the transport equation. This acknowledges that in shear flows and localized vortices, the gradients of the correction term can be significantly larger than those of the smooth equilibrium distribution.
3. Transport Equation Formulation: A simplified collision integral (BGK-type) is used, leading to a kinetic transfer equation for $\varphi$. This equation features a source term dependent on the gradient of the average mass velocity and a sink term representing carrier absorption (inverse free path).
4. Integral Solution: The transport equation is solved by assuming carrier motion is significantly faster than the bulk medium motion. The solution for $\varphi$ is expressed as an integral over carrier trajectories from the medium boundary to the observation point, incorporating an exponential decay factor based on the free path length.
5. Derivation of the Stress Tensor: The viscous stress tensor is calculated by integrating the product of the correction term, the equilibrium distribution, and the momentum flux over the phase space. By transforming the trajectory integration into a volume integration over the medium, the author derives a non-local, integral representation of the stress tensor.

Key Contributions and Results
The primary result is the derivation of a generalized viscous stress tensor $\sigma_{ik}$ that is an integral over spatial coordinates rather than a local differential operator.

• Non-Local Representation: The new expression for $\sigma_{ik}$ (Equation 2.13) explicitly depends on the velocity gradients at all points within the medium, weighted by a kernel involving the distance between points and the free path length. This structure mirrors non-local transport theories, such as the Biberman-Holstein model for resonant radiation transfer.
• Recovery of Standard Theory: In the limit of small path lengths (small gradients), the integral expression reduces to the standard local viscous stress tensor (Equation 1.2), with the viscosity coefficient $\eta = P\tau$. This ensures consistency with established theory in the appropriate regime.
• Resolution of the Paradox: The generalized form resolves the inverse dependence of viscosity on the collision cross-section. The author posits that the paradox arises because the standard derivation assumes small gradients, a condition incompatible with the long-lived localized vortices that act as carriers of disturbances in turbulent flows.
• Conservation Laws: The derivation demonstrates that the correction to the distribution function preserves the total number of particles in the medium, representing a dynamic exchange between the main medium and the ensemble of disturbance carriers.

Significance and Claims
The paper claims that this work constitutes a step toward a non-local theory of hydrodynamic turbulence derived from first principles.

• Modeling Discontinuities: The generalized tensor provides a mathematical operator capable of describing tangential discontinuities and separation flows where velocity derivatives are unbounded, addressing a long-standing gap in numerical modeling.
• Connection to Richardson's Law: The non-local structure of the derived tensor is presented as a necessary foundation for numerically modeling the non-locality of turbulence, specifically to reproduce the empirical Richardson $t^3$ law for pair correlations.
• Validation of Kolmogorov's Approach: The author suggests that this generalization may explain the success of Kolmogorov's dimensional reasoning in predicting turbulence phenomena, despite the fact that these results have historically been difficult to derive directly from the standard Navier-Stokes equation.
• Broader Applicability: The methodology is noted as potentially applicable to heat transfer in gases and liquids, as well as to more complex media like plasma, offering a unified approach to non-local transport phenomena.

The paper concludes that while the derived expression is a significant theoretical advancement, the full demonstration of its ability to reproduce the Richardson law requires subsequent adequate numerical modeling.