🔢 mathematics

Well-posedness and longtime behavior of the conserved Navier--Stokes--Allen--Cahn equation with unmatched viscosities and singular potential

This paper establishes the well-posedness and longtime behavior of the conserved Navier–Stokes–Allen–Cahn equation with unmatched viscosities and singular potentials in both two and three dimensions, proving local strong solution existence, weak solution uniqueness, and the asymptotic convergence of global weak solutions to unique equilibria with phase separation.

Original authors: Maurizio Grasselli, Christoph Hurm, Andrea Poiatti

Published 2026-06-01

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Maurizio Grasselli, Christoph Hurm, Andrea Poiatti

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 you have a glass of liquid containing two types of oil that don't mix well, like water and oil, but they are swirling together in a complex dance. This paper is about a mathematical model that tries to predict exactly how this mixture moves and separates over time, even when the "thickness" (viscosity) of the two oils changes as they mix.

Here is a breakdown of what the authors, Grasselli, Hurm, and Poiatti, achieved, using simple analogies.

The Setup: A Chaotic Dance Floor

The authors are studying a system called the Conserved Navier–Stokes–Allen–Cahn equation.

The Navier–Stokes part is like the rules for how a fluid flows (like wind or water).
The Allen–Cahn part is like a rule for how the two different "oils" decide to separate or mix.
The "Conserved" part means the total amount of each oil stays the same; nothing is created or destroyed, just rearranged.
The "Singular Potential" is a special rule that acts like a bouncer at a club. It says, "You can be almost pure Oil A or almost pure Oil B, but you cannot be exactly 100% one or the other in the middle of the mix." It creates a mathematical "wall" that prevents the mixture from hitting the extreme edges (pure phases) too easily.

The authors are looking at a scenario where the two oils have different thicknesses (viscosities) that change depending on how much of each is present. This makes the math very tricky, like trying to predict the movement of a crowd where some people are wearing heavy boots and others are on roller skates, and the mix of boots and skates changes constantly.

The Main Achievements

1. Proving the Dance Can Start (Local Existence)

The Claim: In a 3D space (our real world), they proved that if you start with a specific, well-behaved mixture, the system has a unique solution for a short period of time.
The Analogy: Imagine you push a complex Rube Goldberg machine. The authors proved that for a short while, the machine will actually work as intended without falling apart immediately, even with the tricky "different thickness" rules. They showed that the math doesn't break down instantly.

2. The "Good Solution" vs. The "Messy Solution" (Uniqueness)

The Claim: They proved that if you have a "strong" solution (a very smooth, well-behaved prediction) and a "weak" solution (a rougher, more general prediction), and they start from the same point, they will stay the same if the rough solution doesn't get too wild.
The Analogy: Think of a GPS route. The "strong solution" is a perfect, high-definition map. The "weak solution" is a sketchy, low-resolution map. The authors showed that as long as the sketchy map doesn't suddenly jump off a cliff (a condition they call "relative energy"), it will follow the exact same path as the high-definition map. If the sketchy map gets too messy, the two might diverge, but they found a way to ensure they stay together under specific conditions.

3. The 2D Special Case

The Claim: In a 2D world (like a flat sheet of paper), they proved that the solution is unique no matter what. There is no need for extra conditions.
The Analogy: In a flat world, the dance is simpler. Even if the dancers are clumsy, there is only one way the dance can end up. The authors closed a gap in previous research that left this question open.

4. The Long-Term Destination (Convergence to Equilibrium)

The Claim: This is the most significant result. They proved that no matter how chaotic the mixture starts, it will eventually calm down and settle into a single, stable state. It won't keep swirling forever.
The Analogy: Imagine a shaken bottle of salad dressing. No matter how violently you shake it, if you leave it alone, the oil and vinegar will eventually separate and settle into a calm, stable layer. The authors proved that this system always settles down, even if the "thickness" of the fluids is just barely continuous (not perfectly smooth). They showed that the mixture eventually stops moving and the phases separate cleanly.

5. The "Magic" of Regularity

The Claim: If the "thickness" of the fluids is a bit smoother (mathematically, Lipschitz continuous), the mixture doesn't just settle; it becomes smoother over time.
The Analogy: If the dancers are wearing slightly better shoes (smoother viscosity), not only do they stop dancing eventually, but their movements become graceful and precise as they slow down. The "rough" edges of the solution smooth out, and the system converges to its final resting state with high precision.

The "Secret Sauce"

The authors used a few clever mathematical tricks to get these results:

Relative Energy: Instead of trying to measure the exact distance between two solutions, they measured the "energy difference" between them. It's like checking if two cars are using the same amount of fuel to see if they are on the same road.
De Giorgi Iterations: A technique used to prove that the mixture stays away from the "pure" extremes (the bouncer's wall) as time goes on. It's like proving that even if you start close to the edge of a cliff, the wind will eventually push you back to the center of the field.
Good Times vs. Bad Times: They realized that the system behaves well most of the time ("good times") and only gets messy occasionally ("bad times"). By focusing on the "good times," they could prove the system eventually settles down.

Summary

In short, this paper takes a very complex, messy fluid dynamics problem involving two mixing fluids with changing thickness and proves three main things:

The math works for a short time.
The predictions are unique (you can't have two different outcomes from the same start) in 2D, and under specific conditions in 3D.
Most importantly: No matter how chaotic the start, the system will eventually calm down, separate cleanly, and settle into a single, stable equilibrium. It proves that the chaos is temporary, and order always wins in the long run.

Technical Summary: Well-Posedness and Longtime Behavior of the Conserved Navier–Stokes–Allen–Cahn Equation with Unmatched Viscosities and Singular Potential

1. Problem Formulation
The paper investigates a mass-conserving model for two-phase incompressible flow with matched densities, described by the Navier–Stokes system nonlinearly coupled with a conserved Allen–Cahn equation. The system is defined on a bounded domain $\Omega \subset \mathbb{R}^d$ ( $d \in \{2, 3\}$ ) and involves the fluid velocity $v$ , the phase-field variable $\phi$ (representing relative concentration), and the chemical potential $\mu$ . The model incorporates:

Unmatched Viscosities: The viscosity coefficient $\nu(\phi)$ depends on the phase field and is assumed to be non-constant.
Singular Potential: The free energy density includes a singular potential $f(\phi)$ , typically of the Flory–Huggins type, which becomes infinite at the pure phases $\phi = \pm 1$ . This enforces the physical constraint $|\phi| < 1$ .
Boundary Conditions: No-slip conditions for velocity ( $v=0$ ) and homogeneous Neumann conditions for the phase field ( $\partial_n \phi = 0$ ).

The governing equations are:
$\begin{aligned} \partial_t v + (v \cdot \nabla)v - \text{div}(\nu(\phi)Dv) + \nabla \pi &= \mu \nabla \phi, \\ \nabla \cdot v &= 0, \\ \partial_t \phi + v \cdot \nabla \phi + \mu - \bar{\mu} &= 0, \\ \mu &= -\Delta \phi + f'(\phi), \end{aligned}$
where $\bar{\mu}$ denotes the spatial average of $\mu$ .

2. Methodology
The authors employ a combination of functional analysis, energy methods, and asymptotic analysis techniques:

Galerkin Approximation: For existence proofs, a full Galerkin scheme is utilized, regularizing the singular potential via Yosida approximation ( $F_\lambda$ ) to handle the singularity.
Energy Estimates: Global weak solutions are established via standard energy inequalities. For strong solutions, higher-order estimates are derived by testing equations with time derivatives and Stokes operators ( $A_S$ ), utilizing Gagliardo–Nirenberg and Agmon inequalities.
Relative Energy Approach: To establish conditional weak-strong uniqueness, the authors utilize a relative energy method. This approach compares a weak solution with a strong solution, controlling the difference via a functional that accounts for the nonlinearity of the potential and the variable viscosity.
De Giorgi Iterations: To prove the asymptotic strict separation property (that the solution stays away from pure phases $\pm 1$ as $t \to \infty$ ), the authors apply De Giorgi's iteration technique to the parabolic equation, adapting it to the coupled system.
Lojasiewicz–Simon Inequality: The convergence to a unique equilibrium is proven using the Lojasiewicz–Simon inequality, which relates the energy functional to the norm of its gradient, provided the potential is real analytic.

3. Key Contributions and Results

Local Well-Posedness in 3D: The paper proves the existence and uniqueness of local-in-time strong solutions in dimension three for initial data with higher regularity ( $v_0 \in H^1_\sigma$ , $\phi_0 \in H^2_N$ , $F'(\phi_0) \in L^2$ ). Crucially, these solutions are shown to be instantaneously strictly separated from the pure phases.
Unconditional Uniqueness in 2D: The authors resolve an open question from previous literature by proving that weak solutions are unconditionally unique in dimension two. This is achieved without requiring the solutions to be strictly separated from pure phases, relying instead on the specific structure of the 2D Ladyzhenskaya inequality.
Conditional Weak-Strong Uniqueness in 3D: A conditional weak-strong uniqueness result is established in dimension three. The condition requires the weak solution to satisfy an integrability condition on the derivative of the potential ( $F'(\phi) \in L^4(0, T; L^4(\Omega))$ ), which is shown to be weaker than the standard assumption of strict separation.
Asymptotic Separation and Convergence:
- Minimal Regularity on Viscosity: The paper demonstrates that any global weak solution asymptotically separates from pure phases and converges to a unique equilibrium in both 2D and 3D. This result holds under minimal assumptions on the viscosity, specifically $\nu \in C^0([-1, 1])$ , without requiring Lipschitz continuity.
- Asymptotic Regularization: If the viscosity is more regular ( $\nu \in W^{1,\infty}(\mathbb{R})$ ), the authors show that any weak solution asymptotically regularizes. This means that after a finite time, the weak solution becomes a strong solution, leading to convergence to the equilibrium in higher-order norms.

4. Significance and Claims
The paper claims several significant advancements over existing literature (specifically referencing works like [15], [18], and [19]):

First-Time Result on Asymptotic Separation: The authors state that this is the first time it is shown that global weak solutions to this coupled system (with unmatched viscosities and singular potentials) asymptotically separate from pure phases and converge to a unique equilibrium in both 2D and 3D, without needing further regularization assumptions initially.
Relaxed Viscosity Assumptions: The convergence to equilibrium is proven even when the viscosity coefficient is merely continuous ( $\nu \in C^0$ ), a significant relaxation compared to previous results that often required Lipschitz continuity.
Resolution of Open Problems: The unconditional uniqueness of weak solutions in 2D, which was left open in prior studies, is rigorously established.
Robustness of the Approach: The methodology combines the asymptotic strict separation property with conditional weak-strong uniqueness to demonstrate that weak solutions eventually regularize if the viscosity is sufficiently smooth, thereby ensuring convergence in higher-order norms.

The work relies on recent seminal results by the first and third authors (referenced as [18, 19]) and extends them to the case of unmatched viscosities and singular potentials, providing a comprehensive picture of the long-time dynamics of the system under minimal structural assumptions.