🔢 mathematics

Weak solutions to the Navier-Stokes equations for steady compressible non-Newtonian fluids

This paper establishes the existence of weak solutions for steady, compressible non-Newtonian Navier-Stokes equations on bounded two- or three-dimensional domains, covering both power-law fluids under specific conditions on the growth exponent and pressure, and Herschel-Bulkley fluids with singular viscosity in a time-discretized setting.

Original authors: Cosmin Burtea, Maja Szlenk

Published 2026-07-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Cosmin Burtea, Maja Szlenk

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where fluids don't just flow like water; they can be thick like honey, stretchy like rubber, or even act like a solid until you push them hard enough. This is the world of non-Newtonian fluids. The paper you provided is a mathematical proof that says: "Yes, we can predict how these tricky fluids will behave in a steady state, even when they are being squished (compressed) and pushed around."

Here is a breakdown of what the authors, Cosmin Burtea and Maja Szlenk, actually did, using simple analogies.

The Big Picture: The "Traffic Jam" of Fluid Particles

Think of a fluid as a massive crowd of people (particles) moving through a city (a bounded room).

Density ( $\varrho$ ): How crowded the room is.
Velocity ( $u$ ): How fast the people are walking.
Pressure ( $p$ ): How much the crowd is pushing against the walls.
Viscosity ( $S$ ): This is the "stickiness" or resistance. In normal water (Newtonian), the stickiness is constant. But in these special fluids, the stickiness changes depending on how fast you try to move. If you move fast, it might get thicker or thinner.

The authors are trying to solve a giant puzzle: Can we prove that a solution exists? In math terms, this means proving that there is at least one valid way to describe the movement of this crowd that satisfies all the physical laws, even if we can't write down a simple formula for it.

The Main Challenge: The "Oscillation" Problem

When mathematicians try to solve these equations, they usually start with a "smoothed out" version of the problem and then try to remove the smoothing to get the real answer.

Imagine trying to take a photo of a fast-moving crowd.

The Approximation: You take a blurry photo (the smoothed version). It's easy to work with.
The Goal: You want to sharpen the photo to see the real people.
The Problem: As you sharpen the photo, the crowd starts to "jitter" or "oscillate." The density of people in one spot might jump wildly between the blurry version and the sharp version.

In the past, for these specific types of fluids, mathematicians couldn't prove that the "jitter" would settle down. They couldn't guarantee that the final, sharp photo actually represented a real physical state. This paper claims to have finally fixed that gap.

The Secret Weapon: The "Energy Balance" Trick

The authors used a clever mathematical trick to prove the solution exists. Think of it like balancing a checkbook.

The Energy Equation: They looked at the total energy of the fluid (how much work is being done to move the crowd).
The Comparison: They compared the energy of their "blurry" approximation against the energy of the "sharp" final result.
The "Singular" Problem: Usually, when you do this math, you run into a division by zero error (like trying to divide by nothing) because the fluid density can be zero in some spots.
The Fix: The authors invented a new way to handle this division. Instead of trying to calculate the whole room at once, they used a "sieve" (a mathematical tool called Egorov's theorem). They said, "Let's ignore the tiny, messy corners of the room where the math is broken, prove the solution works on the clean, big part of the room, and then show that the messy corners don't matter."

By doing this, they proved that the "jitter" in the density and velocity eventually stops. The blurry photo and the sharp photo match up. The solution is stable.

The Specific Fluids They Studied

They focused on two main types of tricky fluids:

Power-Law Fluids: These are fluids where the thickness changes based on speed. If you stir them fast, they might get thinner (like ketchup) or thicker (like cornstarch and water). The authors proved solutions exist for these, provided the "power" of the change isn't too extreme.
Herschel-Bulkley Fluids: These are even stranger. They act like a solid block until you push them hard enough to break them, and then they flow like a liquid. Think of toothpaste: it sits in the tube (solid) until you squeeze it (breaks), then it flows. The authors proved that even for these "solid-then-liquid" fluids, a mathematical solution exists.

What They Did Not Do

It is important to stick to what the paper claims:

No New Applications: They did not design a new machine, a new medical device, or a new industrial process.
No Clinical Uses: They did not study blood flow in patients or how these fluids affect the human body.
No Time-Traveling Solutions: They solved the "steady" case (where the fluid flow looks the same over time, like a river flowing at a constant speed). They did not solve the full, chaotic, changing-in-time version of the problem (though they showed a step toward it).

The Bottom Line

The authors, Burtea and Szlenk, have built a mathematical bridge. Before this paper, there was a gap in our understanding of how to prove these complex, compressible, sticky fluids behave predictably. They filled that gap using a clever "sieve" method to handle the messy parts of the math.

In short: They proved that for a wide range of weird, thick, compressible fluids, the laws of physics don't break down. A solution exists, even if it's too complex to write down on a napkin.

Technical Summary: Weak Solutions to the Steady Navier-Stokes Equations for Compressible Non-Newtonian Fluids

Problem Statement
This paper investigates the existence of weak solutions for the steady Navier-Stokes system governing compressible, barotropic, non-Newtonian fluids on a bounded, $C^2$ domain $\Omega \subset \mathbb{R}^d$ (where $d \in \{2, 3\}$ ). The system is defined by the continuity equation $\text{div}(\varrho u) = 0$ and the momentum equation:
$\text{div}(\varrho u \otimes u) - \text{div} S(Du) + \nabla p(\varrho) = \varrho f + g,$
with no-slip boundary conditions $u=0$ on $\partial\Omega$ . The pressure follows a $\gamma$ -law, $p(\varrho) = \varrho^\gamma$ (with $\gamma > 1$ ), and the viscous stress tensor $S(Du)$ is a nonlinear function of the symmetric gradient $Du$. The authors assume $S$ satisfies monotonicity and power-law growth conditions with exponent $r > 1$ .

The primary mathematical challenge addressed is the existence of solutions when the stress tensor is non-Newtonian (specifically power-law fluids) and the fluid is compressible. In the compressible case, obtaining weak solutions is significantly more difficult than in the incompressible case due to the need to control the oscillations of the density sequence. Previous results for compressible non-Newtonian fluids often relied on structural assumptions ensuring the velocity divergence belongs to $L^\infty$ or that the stress tensor is "almost isotropic." This work aims to remove the "almost isotropic" restriction for power-law fluids.

Methodology
The authors employ a regularization and limit-passage strategy, constructing solutions as limits of approximate systems. The methodology proceeds in three main stages:

Approximate Systems: The authors introduce a regularized system involving parameters $\alpha, \delta, \varepsilon, \eta$ . This system includes artificial viscosity terms ( $-\varepsilon \Delta \varrho$ ), artificial pressure terms, and mollified convective terms ( $\omega_\delta * T_\delta(u)$ ) to ensure the existence of smooth solutions via the Leray-Schauder fixed-point theorem.
Uniform Estimates: A critical part of the proof involves deriving uniform bounds for the approximate solutions $(\varrho_\varepsilon, u_\varepsilon)$ independent of the regularization parameters. The authors establish bounds in $L^{\frac{r\gamma}{r-1}}(\Omega)$ for the density and $W^{1,r}_0(\Omega)$ for the velocity. The proof distinguishes between cases where $r < d$ , $r = d$ , and $r > d$ , utilizing Sobolev embeddings and the Bogovskii operator to test the momentum equation.
Strong Convergence and Identification of Limits: The core difficulty lies in passing to the limit in the nonlinear terms, specifically identifying the weak limits $S(Du)$ and $\varrho^\gamma$ $ϱ^{γ}$ with the nonlinear functions of the weak limits.
- Energy Identity: The authors derive a local energy identity by testing the momentum equation with $\varphi u_\varepsilon$ . By subtracting the energy identity of the limit system, they obtain a relation involving the defect measure $S(Du) : Du - S(Du) : Du$.
- Measure-Theoretic Argument: A key innovation is the handling of the singular term $(\varrho^\gamma)^{\frac{1}{\gamma}-1}$ in the energy identity. Instead of relying on higher integrability estimates for $\nabla u$ (which required "almost isotropic" assumptions in prior work like [7]), the authors apply a measure-theoretic argument based on Egorov's theorem. They prove that for any $\eta > 0$ , there exists a compact set $\Omega_\eta$ where the defect measure vanishes. This allows them to establish the strong convergence of $\nabla u_\varepsilon$ to $\nabla u$ in $L^p(\Omega)$ for $p < r$ .
- Density Convergence: Once the strong convergence of the velocity gradient is established, the strong convergence of the density is recovered by analyzing the momentum equation as a Poisson equation for the pressure, utilizing the strong convergence of the stress tensor and convective terms.

Key Contributions and Results

Existence for Power-Law Fluids: The paper proves the existence of weak solutions for steady compressible non-Newtonian fluids with power-law stress tensors. The main result (Theorem 1.3) holds under the conditions:
- $d > r > \frac{3d}{d+2}$ and $\gamma > \frac{d(r-1)}{(d+2)r - 3d}$ , OR
- $d \leq r$ and $\gamma > 1$ .
  This generalizes previous results by removing the requirement that the fluid be "almost isotropic," a condition previously necessary to control the oscillations of the density.
Stability: Theorem 1.4 establishes the stability of these weak solutions with respect to the source terms $f$ and $g$ . If the source terms converge weakly-*, the corresponding sequence of solutions converges strongly for the density and the velocity gradient (in $L^p$ for $p<r$ ).
Herschel-Bulkley Fluids: The authors extend their results to Herschel-Bulkley fluids, which possess a singular part in their viscosity (viscoplastic behavior). Theorem 1.5 establishes the existence of weak solutions for a time-discretized model of these fluids, where the stress tensor includes a yield stress term $\tau^*$ . This result is presented as a step toward constructing solutions for the non-steady case.

Significance and Claims
The paper claims to resolve the open question of existence for steady compressible non-Newtonian power-law fluids with $\gamma$ -law pressure without imposing "almost isotropic" structural constraints on the stress tensor. The authors explicitly note that while a similar identity (relating energy defects to density oscillations) was proposed by Zhikov and Pastukhova [41] for the non-steady case, a gap in their argument regarding the rigorous justification of the product of measures and functions remained unfilled (as noted in [20]).

The authors' contribution lies in providing a rigorous justification of this identity for the steady case using Egorov's theorem and measure-theoretic localization, thereby bypassing the need for higher integrability of the velocity gradient. They state that their method allows for the recovery of strong convergence for the density and velocity gradient under broader conditions than previously possible. The work is framed as a foundational step, particularly for the Herschel-Bulkley results, toward the more complex non-steady (evolutionary) case, which remains an open problem for the general non-Newtonian setting.