Multi-Sink Solutions to the Self-Similar Euler Equations

✨

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 vast, invisible ocean of fluid (like air or water) that has no friction and never gets squished. In physics, we use the Euler equations to predict how this fluid moves. Usually, if you know how the fluid is moving at the start, you can predict exactly how it will move later. This is called "uniqueness."

However, mathematicians have a nagging suspicion: What if the rules break down? What if, starting from the exact same initial conditions, the fluid could suddenly decide to swirl in two completely different ways? This is the "Yudovich Problem," a famous mystery in fluid dynamics.

This paper by Choi and Coiculescu is like a detective story trying to find a "smoking gun" that proves this non-uniqueness is possible. They do this by looking for a very special, strange type of fluid motion called a Multi-Sink Solution.

Here is the breakdown of their discovery using simple analogies:

1. The "Self-Similar" Zoom

Imagine you have a picture of a storm. If you zoom in on the center of the storm, it looks exactly like the whole storm, just smaller. If you zoom in again, it still looks the same. This is called self-similarity.

The authors are looking for fluid patterns that look the same no matter how much you zoom in or out. These are "homogeneous" solutions. They are the building blocks of the fluid's behavior.

2. The "Stagnation Point" (The Dead Center)

In a swirling fluid, there are usually spots where the water isn't moving at all. We call these stagnation points.

The Old Rule: For a long time, mathematicians thought that if you had a self-similar fluid pattern, there could only be one dead center (a single stagnation point), located right in the middle of the swirl.
The New Discovery: This paper proves that you can actually build a fluid pattern with multiple dead centers (multiple stagnation points) away from the middle.

3. The "Gluing" Analogy (How they built it)

How did they find these multi-center patterns? They didn't just guess; they glued pieces together.

Imagine you have a set of Lego bricks. Each brick is a valid, smooth piece of fluid motion that works perfectly on its own, but only in a specific wedge-shaped slice of the circle (like a slice of pizza).

The Problem: If you just put them side-by-side, the edges might not match up perfectly. The fluid might tear or jump.
The Solution: The authors found specific "bricks" (mathematical solutions) that fit together like a puzzle. When they glued two different types of bricks together, they created a new, larger pattern.
The Result: By gluing these pieces in a specific alternating pattern (Positive, Negative, Positive, Negative), they created a fluid flow that has two distinct "sinks" (places where the fluid drains inward) away from the center, plus a "saddle" point in the very middle.

Think of it like a dance floor. Usually, everyone dances around one central DJ booth. The authors figured out how to arrange the dancers so there are two distinct groups pulling people toward two different spots on the floor, while the center remains a chaotic crossing point.

4. The "Rough Edge" (Why it's weird)

Here is the catch: These new multi-sink solutions aren't perfectly smooth.

The Smooth World: In the ideal world, fluid flows are smooth and continuous.
The Rough Reality: Because the authors had to "glue" these pieces together, the resulting fluid has cusps (sharp corners) along the lines where the pieces meet. The speed of the fluid changes abruptly there.
The Significance: This "roughness" is actually a feature, not a bug. It proves that if you have more than one stagnation point, the fluid must be rough and discontinuous. This breaks the "smoothness" assumption that usually guarantees a unique solution.

5. The "Two-Sink" Solution and the Big Picture

The authors highlight a specific pattern called the "Two-Sink Solution."

It looks like a butterfly or a figure-eight.
As they tweak the math (changing a parameter called $\alpha$ ), this two-sink pattern slowly morphs into a very simple, well-known flow called a "shear flow" (where layers of fluid slide past each other).
Why this matters: This suggests a "bifurcation" (a fork in the road). It implies that from a simple starting point, the fluid could potentially split into two different futures: one simple flow, or one complex, multi-sink flow.

The Takeaway

This paper is a major step toward proving that the Euler equations might not always have a single, unique answer.

Before: We thought fluid motion was like a train on a single track.
Now: This paper shows that under certain extreme conditions, the track might split. The fluid could choose to go down a path with two swirling centers instead of one.

While they haven't solved the entire mystery of fluid uniqueness yet, they have built the first concrete "bridge" (the multi-sink solution) showing that the strange, non-unique behavior is mathematically possible. They found the "ghost" in the machine that suggests the fluid might have a mind of its own.

1. Problem Statement and Motivation

The paper addresses the non-uniqueness problem for the two-dimensional incompressible Euler equations. Specifically, it investigates whether weak solutions with unbounded but integrable vorticity (the "Yudovich problem") are unique.

Context: While uniqueness is known for bounded vorticity, it remains an open question for $L^1 \cap L^p$ vorticity. A promising strategy to prove non-uniqueness is the construction of self-similar solutions that bifurcate from homogeneous initial data.
The Gap: Previous constructions of self-similar solutions (spirals) relied on small perturbations of power-law vortices. These solutions possess a single stagnation point in their pseudo-velocity field.
The Hypothesis: The authors propose that a mechanism for non-uniqueness requires the existence of self-similar profiles where the pseudo-velocity field possesses multiple stagnation points (specifically, multiple "sinks" away from the origin).
Goal: To construct and classify homogeneous self-similar solutions to the 2D Euler equations that exhibit these multi-sink properties.

2. Methodology

The authors employ a combination of Hamiltonian dynamical systems analysis, asymptotic analysis of elliptic integrals, and gluing techniques.

A. Reduction to an ODE System

The authors consider homogeneous solutions of the form $u = \nabla^\perp \Psi$ where $\Psi = r^\lambda \psi(\theta)$ . Substituting this into the steady Euler equations reduces the problem to a nonlinear ordinary differential equation (ODE) for the angular profile $\psi(\theta)$ :
$-\lambda \psi \psi'' + (\lambda - 1)(\psi')^2 - \lambda^2 \psi^2 + 2(\lambda - 1)P = 0$
where $P$ is a constant related to pressure. This system is identified as a Hamiltonian system with a first integral (Bernoulli function).

B. Local Solutions and Phase Portrait

The authors analyze the phase portrait of the ODE in the $(\psi, \psi')$ plane. They identify two types of maximal local solutions depending on the sign of a constant $B$ (derived from the Bernoulli function):

$\psi_+$ : Solutions corresponding to $B=1$ .
$\psi_-$ : Solutions corresponding to $B=-1$ .
These solutions are defined on finite intervals (lifespans) $T_+$ and $T_-$ , which depend on the pressure parameter $P$ .

C. The Gluing Procedure

To construct a global $2\pi$ -periodic solution, the authors "glue" these local solutions together at their endpoints (where $\psi=0$ ).

The gluing is performed by alternating signs (e.g., $\psi_+, -\psi_-, \psi_+, \dots$ ).
The resulting function $\psi(\theta)$ is $C^1$ but generally not $C^2$ (specifically $C^{1, 2-2/\lambda}$ ), leading to discontinuous vorticity along the rays where the gluing occurs.
Condition for Existence: A valid global solution exists if the sum of the lifespans of the glued segments equals $2\pi$ . This requires solving for a specific pressure $P < 0$ such that:
$\sum T_i(P) = 2\pi$

D. Asymptotic Analysis

To prove the existence of such a $P$ , the authors perform a rigorous asymptotic analysis of the "transcendental elliptic integrals" $T_+(P)$ and $T_-(P)$ as $P \to 0^-$ . They utilize Mellin transforms and expansion techniques to determine the monotonicity of the sum of lifespans with respect to $P$ .

3. Key Contributions and Results

A. Classification of Multi-Sink Solutions (Theorem 3.3)

For the scaling parameter range $\lambda \in (1, 2)$ (equivalent to self-similar exponent $\alpha \in (0, 1)$ ), the authors provide a complete classification of all possible $2\pi$ -periodic solutions obtained by gluing. The valid combinations of local solutions ( $\psi_+$ and $\psi_-$ ) are:

$2k$ copies of $\psi_-$ ( $k \ge 2$ ).
One $\psi_+$ and $2k-1$ copies of $\psi_-$ ( $k \ge 2$ ).
Two copies of $\psi_+$ (This corresponds to the shear flow).
Two copies of $\psi_+$ and two copies of $\psi_-$ .
Two copies of $\psi_+$ and $2k$ copies of $\psi_-$ ( $k \ge 2$ , valid for $\lambda \le 3/2$ ).

B. Existence of the Two-Sink Solution

The paper highlights a specific solution formed by gluing two $\psi_+$ and two $\psi_-$ segments in the sequence $\psi_+, -\psi_-, \psi_+, -\psi_-$ .

Symmetry: This solution is 2-fold rotationally symmetric.
Stagnation Points: The pseudo-velocity field $U - \frac{\xi}{\alpha}$ $U - \frac{ξ}{α}$ possesses:
- A saddle point at the origin.
- Two sink stagnation points located away from the origin.
Regularity: The solution is $C^{1, 2-2/\lambda}$ , meaning the velocity is Hölder continuous, but the vorticity is discontinuous and unbounded along the rays connecting the origin to the sinks.

C. Non-Uniqueness Implications

Proposition 3.6: The authors prove that any homogeneous self-similar solution with more than one stagnation point must have discontinuous vorticity. This confirms that multi-sink solutions are inherently irregular, distinguishing them from classical smooth solutions.
Bifurcation from Shear Flow: The two-sink solution is shown to bifurcate from the power-law shear flow as the scaling parameter $\alpha \to 0^+$ $α \to 0^{+}$ (or $\lambda \to 2^-$ $λ \to 2^{-}$ ).
- As $\lambda \to 2$ , the critical pressure $P^*$ required to sustain the two-sink structure vanishes quadratically ( $|P^*| \sim \alpha^2$ ).
- The solution converges to the power-law shear flow in the $C^{1,\gamma}$ norm.

D. Numerical Validation

The authors provide numerical evidence supporting the asymptotic analysis, specifically verifying the convergence rate of the critical pressure $P^*$ to zero as $\lambda \to 2$ . Vector plots of the pseudo-velocity field confirm the presence of the two distinct sinks.

4. Significance

Advancement of the Non-Uniqueness Program: This work provides the first rigorous construction of self-similar Euler solutions with multiple stagnation points. This is a crucial step toward the non-uniqueness scenario proposed by Bressan and others, which relies on the existence of such multi-sink profiles to create bifurcations in the solution space.
Completeness of Classification: For the range $\alpha \in [1/2, 1)$ , the paper completes the classification of all homogeneous steady states of the 2D Euler equations, distinguishing between those obtained by gluing (multi-sink) and those that are not.
Regularity Insights: The paper establishes a direct link between the topology of the flow (number of stagnation points) and the regularity of the solution. It proves that multi-sink structures necessitate vorticity singularities, offering a new perspective on the "Yudovich problem" boundary.
Connection to Shear Flows: By demonstrating the bifurcation of the two-sink solution from the shear flow, the authors bridge the gap between simple homogeneous flows and complex spiral structures, suggesting a pathway for how non-uniqueness might emerge from generic initial data.

In summary, Choi and Coiculescu successfully construct a new class of "multi-sink" self-similar solutions, proving their existence, classifying their possible configurations, and demonstrating their relevance to the fundamental question of uniqueness in fluid dynamics.