🔢 mathematics

Non-Smooth Solutions of the Navier-Stokes Equation and their Means

This paper constructs non-smooth, finite-time blowup Leray-Hopf solutions to the incompressible Navier-Stokes equations on a periodic 3D torus driven by turbulent fluctuations, while demonstrating that the mean value of such weak solutions corresponds to a smooth solution.

Original authors: J. Glimm, J. Petrillo

Published 2026-07-07

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: J. Glimm, J. Petrillo

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

The Big Picture: A Battle Between Order and Chaos

Imagine the Navier-Stokes equation as the ultimate rulebook for how fluids (like water or air) move. For over a century, mathematicians have been trying to answer a specific question: If you start with a smooth, calm fluid, will it stay smooth forever, or can it suddenly "break" into chaos?

This paper claims to have found a specific scenario where the fluid does break. The authors argue that if you start with a fluid that has a specific type of "turbulent jitter" (energy fluctuations), the math predicts that the smooth flow will eventually hit a wall and explode into a singularity (a "blowup") in a finite amount of time.

However, there is a twist: If you take the average of all these chaotic movements, that average remains perfectly smooth and calm forever.

The Two Characters: The "Mean" and the "Fluctuation"

To understand the paper, think of a crowd of people in a stadium doing "The Wave."

The Mean (The Smooth Solution): This is the average movement of the crowd. If you look at the crowd from a helicopter, you see a smooth, rolling wave moving around the stadium. The authors prove that this "average wave" is always smooth, predictable, and never breaks. It follows a simple, calm path (mathematically, it acts like a heat equation).
The Fluctuation (The Turbulent Solution): This is the individual jitter of every single person in the crowd. Some are jumping, some are shuffling, some are standing still. The paper focuses on a specific type of crowd where these individual jitters are energetic and chaotic.

The Conflict: Why the "Break" Happens

The paper sets up a logical trap involving three facts about how energy dissipates (fades away) in a fluid:

Fact A (The Smooth Path): The overall fluid flow (the "Mean") is supposed to fade away at a certain steady, slow speed, like a cup of coffee cooling down.
Fact B (The Chaotic Path): The individual turbulent jitters (the "Fluctuations") are also fading away, but the authors argue they fade away much faster than the smooth flow.
Fact C (The Contradiction): Physics demands that the chaotic jitters must always be larger than the smooth flow they are part of. You can't have a crowd where the individual people are jumping less than the average wave they are creating.

The Analogy:
Imagine a runner (the smooth flow) and a swarm of bees (the turbulent fluctuations) flying around them.

The runner slows down at a steady pace.
The bees are supposed to slow down even faster.
But the bees must always be buzzing faster than the runner is moving.

Eventually, the bees slow down so much that they would have to stop buzzing entirely to keep up with the runner's speed. But the math says they must keep buzzing. This creates a logical impossibility. The paper argues that the only way to resolve this contradiction is for the smooth runner to suddenly trip and fall. In math terms, the solution "blows up" (becomes non-smooth) at a specific time, $T^*$ .

The "Entropy Principle" (The Rule of Maximum Chaos)

The authors use a concept called the Entropy Principle. Think of entropy as a measure of "messiness" or "disorder."

The paper assumes the fluid behaves in a way that maximizes this messiness. It chooses the path of maximum chaos.
Under this rule of maximum chaos, the turbulent jitters are so aggressive that they force the smooth flow to break down.

The "Mean" Saves the Day

While the individual chaotic path leads to a crash (blowup), the paper makes a second major claim: The Average is Safe.

If you take all the possible chaotic paths the fluid could take and average them out, the result is a "Mean" solution.

This "Mean" solution has zero entropy (it is perfectly ordered).
Because it is perfectly ordered, it does not suffer from the contradiction described above.
The "Mean" never blows up. It stays smooth forever.

The Millennium Prize Connection

There is a famous million-dollar math prize (the Millennium Prize) asking if smooth fluids always stay smooth.

The Paper's Verdict: The authors claim the answer is NO. They say that if you start with "turbulent" initial data (specifically, data with non-zero energy fluctuations), the fluid will eventually break.
The Caveat: They admit that if you start with "smooth" data (the "Mean"), it stays smooth. But since the prize problem asks if any smooth start leads to a smooth finish, and they found a start that leads to a break, they claim to have solved the problem in the negative.

Summary of the Claims

Blowup Exists: If you start a fluid with specific turbulent energy fluctuations, the math predicts it will become non-smooth (blow up) in a finite time.
The Reason: It's a race between the smooth flow and the turbulent jitters. The jitters fade too fast to stay consistent with the smooth flow, forcing a crash.
The Average is Safe: The "average" of these chaotic fluids is always smooth and never breaks.
The Method: They prove this by working in a specific mathematical "space" (called $V^*$ ) that allows them to see these contradictions clearly, using the concept of maximizing entropy (messiness).

In short: The paper argues that while the average behavior of a fluid is calm and predictable, the actual behavior of a turbulent fluid can suddenly snap and break, proving that smoothness is not guaranteed for all starting conditions.

Technical Summary: Non-Smooth Solutions of the Navier-Stokes Equation and their Means

Problem Statement
This paper addresses the regularity of solutions to the incompressible Navier-Stokes equations in a periodic spatial domain $T^3$ . Specifically, it investigates the existence of non-smooth (finite-time blowup) Leray-Hopf (LH) weak solutions and characterizes the relationship between these singular solutions and their "mean" values. The work positions itself as a response to the Clay Mathematics Institute's Millennium Prize problem regarding the existence and smoothness of solutions in three dimensions. The authors claim to solve the problem in the negative by constructing initial data that leads to finite-time blowup, while simultaneously demonstrating that the mean of any weak solution remains smooth for all time.

Methodology
The analysis is conducted within the framework of weak solutions, utilizing specific function spaces and entropy principles:

Function Spaces: The authors work primarily in the dual space $V^*$ (where $V=H^1$ ) rather than the standard energy space $H=L^2$ . This choice simplifies proofs regarding analyticity and blowup. The solutions are analyzed in the space $W = L^p(0, T; V^*)$ with specific cutoff functions $\phi \in L^1(0, T; V)$ .
Entropy Principle: A central methodological tool is the "Entropy Principle," which posits the existence of weak solutions that maximize entropy production. The paper distinguishes between "entropy production maximizing" solutions (associated with turbulent fluctuations) and "entropy production minimizing" solutions (associated with the mean).
Decay Rate Analysis: The proof of blowup relies on comparing decay rates of different components of the solution. The authors analyze the decay of the solution $u$ $u$ versus the decay of turbulent fluctuation integrands $\nu_t^{int}$ $ν_{t}^{in t}$ .
- The solution $u$ (in the context of the SRI, or Summation over Repeated Indices, subspace) exhibits a slow viscous decay rate of $O(e^{-\nu t/2})$ .
- The turbulent fluctuations $\nu_t^{int}$ (in the NSRI, or Non-Summed Repeated Indices, subspace) are shown to decay at a faster rate, $O(e^{-4\nu t/3})$ (or $O(t^{-4/3})$ in polynomial terms).
Cutoffs and Limits: The construction utilizes spatial and temporal cutoffs (finite-dimensional subspaces of smooth basis elements) to define weak solutions. The results are established as limits where these cutoffs are removed.
Feynman Diagrams: The blowup mechanism is described in terms of connected Feynman diagrams, where a discontinuity occurs in specific non-SRI expansion coefficients (specifically those related to turbulent dissipation) as they transition from turbulent to viscous modes.

Key Contributions and Results

Construction of Finite-Time Blowup:
The paper constructs non-smooth Leray-Hopf weak solutions that experience finite-time blowup.
- Initial Data: The blowup is triggered by "turbulent" initial data, characterized by non-zero turbulent fluctuations in the energy analysis space (specifically, non-zero NSRI energy modes).
- Mechanism: The blowup arises from a logical contradiction in the decay rates for entropy-maximizing solutions. The turbulent fluctuation $\nu_t^{int}$ decays faster than the solution $u$ itself. Eventually, at a finite time $T^*$ , the fluctuation decay curve falls below the lower bound of the viscous decay allowed for the smooth solution. To resolve this contradiction, the solution must lose smoothness, resulting in a discontinuity in the turbulent dissipation terms.
- Result: Theorem 5.1 establishes that an entropy-maximizing solution with turbulent initial conditions has only a finite interval of regularity and is not globally smooth.
Regularity of the Mean:
The paper identifies the "mean" of a weak solution as a distinct entity with different regularity properties.
- Definition: The mean $M(u)$ is defined in the vector Fourier space as a sum of means over configuration space, phase space, and fixed time.
- Entropy Minimization: The mean corresponds to an entropy production minimizing solution. For this class of solutions, the turbulent dissipation rate $\nu_t$ is zero.
- Smoothness: Theorem 7.1 proves that the mean of any weak solution is a smooth solution for all positive times.
- Equation Reduction: For the mean, the nonlinear terms and turbulent fluctuations vanish ( $M(\nu_t^{NL}) = 0$ and $M(\nu_t^{temp}) = 0$ ). Consequently, the evolution equation for the mean reduces to the heat equation: $\partial_t M(u) = \nu \Delta M(u)$ . Since solutions to the heat equation with $H$ initial data are smooth for $t > 0$ , the mean is smooth.
Resolution of the Millennium Prize Problem:
The authors claim to solve the Millennium Prize problem in the negative.
- They demonstrate that for certain turbulent initial conditions (which are valid within the problem's constraints of smooth, divergence-free, periodic data), the solution blows up in finite time.
- They clarify that while the mean of a solution is always smooth, the mean does not possess "general" initial data (it is a specific subset derived from the weak solution). Therefore, the smoothness of the mean does not constitute a positive solution to the prize problem, which requires smoothness for all smooth initial data.

Significance and Claims
The paper claims to provide a constructive proof of the existence of non-smooth solutions to the 3D incompressible Navier-Stokes equations, thereby refuting the conjecture that smooth solutions exist globally for all smooth initial data.

Negative Solution: By showing that turbulent initial data leads to finite-time blowup, the paper asserts it has solved the Millennium Prize problem in the negative.
Distinction between Turbulence and Mean: A significant conceptual contribution is the rigorous separation of the "turbulent" component (which blows up) from the "mean" component (which remains smooth and satisfies a heat equation). The paper argues that the mean value of a weak solution is always a smooth solution, but this does not prevent the underlying weak solution from becoming singular.
Role of Entropy: The work emphasizes the role of entropy production maximization in driving the blowup of turbulent fluctuations, contrasting this with entropy minimization which yields the smooth mean.

The authors note that their analysis is restricted to incompressible, isothermal, single fluids in a periodic domain. They rely on the existence of entropy production maximizing solutions (established in prior literature) and extend these results to demonstrate the specific decay rate contradictions that necessitate blowup. The paper does not propose experimental verification or future applications but frames its contribution as a theoretical resolution to the regularity question.