Abstract
We are concerned with the inviscid limit of the Navier-Stokes equations on bounded regular domains in $\mathbb{R}^3$ with the kinematic and Navier boundary conditions. We first establish the existence and uniqueness of strong solutions in the class $C([0,T_\star); H^r(Ω; \mathbb{R}^3)) \cap C^1([0,T_\star); H^{r-2}(Ω;\mathbb{R}^3))$ for $T_\star > 0$ for the initial-boundary value problem with divergence-free initial data in the Sobolev space $H^r(Ω; \mathbb{R}^3)$ for $r \geq 2$. For strong solutions with $H^{r+1}$ regularity, we establish the inviscid limit in $H^r(Ω; \mathbb{R}^3)$ uniformly on $[0,T_\star)$ for $r > 5/2$. This demonstrates that boundary layers do not develop up to the highest order Sobolev norm in $H^r(Ω; \mathbb{R}^3)$. Furthermore, we present an intrinsic geometric proof for the failure of the strong inviscid limit under a non-Navier slip-type boundary condition.
Introduction
A fundamental question in fluid mechanics is the behavior of a viscous fluid as its viscosity vanishes. This transition is described by the "inviscid limit," where the solutions to the Navier-Stokes equations (governing viscous flow) are expected to converge to the solutions of the Euler equations (governing ideal, frictionless flow).
While this limit is well-understood in whole-space domains, it remains a significant challenge in bounded domains. The primary obstacle is the "no-slip" boundary condition, where fluid adheres to the container walls. This condition typically triggers the formation of a boundary layer—a region of intense velocity gradients near the wall—which prevents convergence in high-order Sobolev norms. This paper investigates whether the Navier boundary condition, which allows for fluid slip with resistance, provides a more regular transition for fluids in curved 3D domains.
Methodology and Geometric Formulation
The study employs a sophisticated framework combining functional analysis with differential geometry to address the complexities of curved boundaries.
Geometric Intrinsic Approach
Rather than relying on local coordinate systems, which can obscure the underlying physics on curved surfaces, the authors reformulate the Navier boundary conditions using intrinsic geometric quantities. Key to this is the shape operator (the second fundamental form), which describes the surface curvature. By incorporating the shape operator directly into the boundary conditions, the authors ensure the mathematical model accounts for how the domain's geometry influences fluid friction.
Analytical Tools
To establish regularity, the authors derive a variant of the Caldéron-Zygmund estimate tailored for these boundary conditions. This allows them to prove that the high-order Sobolev norm ($H^{r+1}$) of the velocity field is equivalent to the $L^2$ norms of its iterated curls (vorticity). Using induction on the regularity index $r$, they establish "a priori" energy estimates that remain bounded even as the viscosity coefficient $\nu$ approaches zero.
Key Results
1. Existence and Strong Convergence
The authors prove that for $r \geq 2$, strong solutions to the Navier-Stokes equations with Navier boundary conditions exist and are unique. More importantly, they demonstrate that as $\nu \to 0$, the viscous solution $u^\nu$ converges strongly to the Euler solution $u$ in the $H^r$ norm for $r > 5/2$. This result is significant because it shows that, unlike the no-slip case, the Navier condition suppresses the development of boundary layers in high-order norms.
2. Quantitative Convergence Rates
The paper establishes a quantitative rate of convergence of $O(\nu^{1/3})$. This rate can be further refined to $O(\sqrt{\nu})$ under the assumption that the velocity gradients remain uniformly bounded, providing a precise measure of how quickly the viscous fluid begins to behave like an ideal one.
3. The Failure of "Simplified" Slip Conditions
A major contribution of this work is the geometric proof regarding the failure of the strong inviscid limit under certain "non-Navier" slip conditions, such as $\omega \times n = 0$ (where $\omega$ is vorticity and $n$ is the normal vector). Using the Lie bracket of vector fields, the authors show that on any domain with non-vanishing Gauss curvature, these simplified conditions force a behavior in the vorticity that is fundamentally incompatible with the Euler equations. This highlights that the true Navier condition is uniquely suited to the geometry of curved walls.
Discussion and Limitations
While the paper provides a robust proof of convergence, it acknowledges certain limitations:
- Temporal Constraints: The results are valid for the lifespan of the strong Euler solution $[0, T_\star)$. Because 3D Euler solutions may potentially experience blow-up, the results are "local in time."
- The Highest-Order Gap: The authors prove convergence in $H^r$ assuming $H^{r+1}$ regularity. Whether the limit holds in the highest possible Sobolev norm ($H^{r+1}$ itself) remains an open mathematical challenge.
- Slip Length Scaling: The study assumes a constant slip length. Future research may explore how the limit behaves if the slip length scales in proportion to the viscosity, a scenario relevant to various turbulence models.
Conclusion
This research clarifies the relationship between boundary conditions and the inviscid limit in 3D curved domains. By demonstrating that Navier boundary conditions prevent the formation of high-order boundary layers, the authors provide a rigorous mathematical foundation for the use of slip conditions in modeling high-Reynolds-number flows where curvature plays a critical role.
Summary updated