On the absence of anomalous dissipation for the Navier-Stokes equations with Navier boundary conditions: a sufficient condition
Claude Bardos, Daniel W. Boutros, Edriss S. Titi · 2025
Read On the absence of anomalous dissipation for the Navier-Stokes equations with Navier boundary conditions: a sufficient condition by Claude Bardos, Daniel W. Boutros, Edriss S. Titi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-12432.txt.
Introduction
Fluids can, in principle, waste energy at infinitely fine scales. That worry is called anomalous dissipation. Claude Bardos and two co-authors give a sufficient condition ruling it out in 2025. Their setting is the three-dimensional Navier-Stokes equations in a bounded region. The boundary uses Navier conditions, where the fluid slips along the wall under friction.
Earlier arguments leaned on extra assumptions. Some concerned the pressure near the boundary. Others needed a strong solution of the inviscid equations at hand. Their proof avoids both. It draws instead on their recent regularity results for the pressure of weak Euler solutions. The condition clarifies when viscosity alone accounts for all energy loss.
Catalog record
| CREATOR | Claude Bardos, Daniel W. Boutros, Edriss S. Titi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-12432
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On the absence of anomalous dissipation for the Navier-Stokes equations with Navier boundary conditions: a sufficient condition Claude Bardos, Daniel W. Boutros, Edriss S. Titi Abstract . We consider the three-dimensional incompressible Navier-Stokes equations in a bounded domain with Navier boundary conditions. We provide a sufficient condition for the absence of anomalous energy dissipation without making assumptions on the behaviour of the corresponding pressure near the boundary or the existence of a strong solution to the incompressible Euler equations with the same initial data. We establish our result by using our recent regularity results for the pressure corresponding to weak solutions of the incompressible Euler equations [Arch. Ration. Mech. Anal., 249 (2025), 28]. 2020 Mathematics Subject Classification 35Q30 (primary), 35D30, 35Q31, 35Q35, 76D05, 76D10 (secondary) This paper is dedicated to Professor Peter Constantin, on the occasion of his 75th birthday, as a token of friendship and admiration for his contributions to research in partial differential equations and fluid mechanics. 1. Introduction We consider the incompressible three-dimensional Navier-Stokes equations in a bounded domain \Omega with \partial\Omega\in C^{4} , which, for a given viscosity \nu>0 , are described by the system (1) \partial_{t}u_{\nu}+\nabla\cdot(u_{\nu}\otimes u_{\nu})-\nu\Delta u_{\nu}+\nabla p_{\nu}=0,\hskip 10.00002pt\nabla\cdot u_{\nu}=0, where u_{\nu}:\Omega\times[0,T]\rightarrow\mathbb{R}^{3} is the velocity field and p_{\nu}:\Omega\times[0,T]\rightarrow\mathbb{R} is the pressure, which are the unknowns. Moreover, system ( 1 ) is supplemented by the following initial data (2) u_{\nu}\lvert_{t=0}=u_{\nu,0}, where u_{\nu,0}\in L^{2}(\Omega) , with \nabla\cdot u_{\nu,0}=0 (in a weak sense). We define n to be the outward normal vector to the boundary \partial\Omega . On the boundary and for a given well-defined vector field v\in C^{0,\alpha}(\Omega) (for \alpha\in(0,1) ), we will denote its projection onto the tangential components as follows (3) (v)_{\tau}\coloneqq\big[v-(v\cdot n)n\big]_{\partial\Omega}. In addition, we will supplement system ( 1 ) with Navier boundary conditions, which consist of a no-flux (impermeability) boundary condition and a complementary boundary condition C (defined in equation ( 5 ), below) (4) u_{\nu}\cdot n=0\hskip 10.00002pt\text{ and }\hskip 10.00002ptC(u_{\nu})=0\hskip 10.00002pt\hbox{ on }\partial\Omega. Similarly to [ 8 , 23 ] , we will consider a generalised (tangential) Navier boundary condition C of the type (5) C(u_{\nu})\coloneqq\left(D(u_{\nu})\cdot n\right)_{\tau}+\lambda(\nu,x)\cdot u_{\nu}=0\hskip 10.00002pt\hbox{ on }\partial\Omega, where D(v)\coloneqq\frac{1}{2}[\nabla v+(\nabla v)^{T}] denotes the symmetric part of the gradient of the vector field v , (for the sake of clarity) we observe that D(u_{\nu})\cdot n denotes the (usual) product between a matrix and a vector field, the tensor-valued function \lambda:(0,\infty)\times\Omega\rightarrow\mathbb{R}^{3\times 3} is assumed to be a given continuous and nonnegative semidefinite function. This class of boundary conditions includes the Robin boundary conditions or boundary conditions involving the vorticity or the stress tensor (in the latter case see, e.g., [ 10 ] ). Now we recall the identity (for v\in H^{2}(\Omega) with (v\cdot n)\lvert_{\partial\Omega}=0 ) (6) (2D(v)\cdot n-(\nabla\times v)\times n)_{\tau}=-2(D(n)\cdot v)_{\tau}\hskip 10.00002pt\hbox{ on }\partial\Omega, which holds in H^{1/2}(\partial\Omega) in the sense of trace on the boundary. The proof of identity ( 6 ) can for example be found in [ 42 , Lemma 3.10] , see also [ 23 , Lemma B.1] . Therefore, thanks to ( 6 ) we can rewrite the Navier boundary condition ( 5 ) as follows (7) (\omega_{\nu}\times n)_{\tau}+2\lambda(\nu,x)\cdot u_{\nu}-2\big(D(n)\cdot u_{\nu}\big)_{\tau}=0\hskip 10.00002pt\hbox{ on }\partial\Omega. If one chooses \lambda=D(n) (which is not necessarily nonnegative semidefinite), the Navier boundary condition ( 5 ) is equivalent to the following vorticity boundary condition (in three dimensions) (8) \omega_{\nu}\times n\lvert_{\partial\Omega}=0. In the case of two dimensions one has instead of ( 6 ) the following identity (9) (D(u_{\nu})\cdot n)\cdot\tau=\frac{1}{2}\omega_{\nu}-\kappa(u_{\nu}\cdot\tau), where \kappa is the curvature of the boundary \partial\Omega . The proof of identity ( 9 ) can be found in [ 13 , Lemma 2.1] , as well as [ 27 , Collorary 4.2] . Observe that in the two-dimensional case the function \lambda is scalar-valued, therefore if we choose \lambda=\kappa in ( 5 ) together with ( 9 ) we end up with the following boundary condition (in two dimensions) (10) \omega_{\nu}\lvert_{\partial\Omega}=0. Formally, if we take the vanishing viscosity limit \nu\rightarrow 0 of the Navier-Stokes equations ( 1 ) we obtain the Euler equations (11) \partial_{t}u+\nabla\cdot(u\otimes u)+\nabla p=0,\hskip 10.00002pt\nabla\cdot u=0, which is supplemented by the following impermeability boundary condition (12) (u\cdot n)\lvert_{\partial\Omega}=0. The dissipation anomaly (also known as the Kolmogorov zeroth law of turbulence) is one of the most fundamental properties of turbulent flows and is a crucial feature of the theory of turbulence. In this context, we assert that the Navier-Stokes equations exhibit anomalous dissipation if there is a sequence of Leray-Hopf weak solutions \{u_{\nu_{k}}\} of the Navier-Stokes equations (for a sequence of viscosities \nu_{k}\rightarrow 0 ) such that (cf. Definition 1 ) \displaystyle\liminf_{\nu_{k}\rightarrow 0}\bigg[ \displaystyle 2\nu_{k}\int_{0}^{T}\lVert D(u_{\nu_{k}})(\cdot,t)\rVert_{L^{2}(\Omega)}^{2}dt \displaystyle+ \displaystyle 2\nu_{k}\sum_{i,j=1}^{3}\int_{0}^{T}\int_{\partial\Omega}\lambda_{ij}(\nu_{k},x)(u_{\nu_{k}})_{i}(u_{\nu_{k}})_{j}dxdt\bigg]>0. The purpose of this contribution is to establish a sufficient condition for the absence of anomalous dissipation in the case of Navier boundary conditions, with no regularity assumptions on the corresponding pressure (near the boundary). In particular, in order to prove this result we employ recently established regularity results for the pressure for Hölder continuous weak solutions of the Euler equations (cf. [ 4 ] ). To be more precise, in [ 4 ] it was shown that Hölder regularity propagates from the velocity field to the pressure for weak solutions of the three-dimensional incompressible Euler equations, specifically if the velocity field of a weak solution u\in L^{\infty}((0,T);C^{0,\alpha}(\Omega)) , for \alpha\in(0,1) , then the corresponding pressure p\in L^{\infty}((0,T);C^{0,\alpha}(\Omega)) (after the same result was established in 2D in [ 9 ] , see also [ 18 , 19 , 30 ] ). We will utilise the results from [ 4 ] in this paper to show a sufficient condition for the convergence of weak solutions of the Navier-Stokes equations to a weak solution of the incompressible Euler equations and the absence of anomalous energy dissipation. Most importantly, we observe that in [ 4 ] it was established that the boundary condition for the pressure for weak solutions of the Euler equations, which is rigorously derived in [ 4 ] from the weak formulation, is different from the standard classical one for strong solutions. The nonequivalence is further demonstrated by means of an explicitly constructed example. The literature regarding the dissipation anomaly (and the associated Onsager conjecture) as well as the vanishing viscosity limit is very substantial. Consequently, the following overview will therefore be necessarily incomplete in reviewing all the relevant work. In the case of no-slip Dirichlet boundary conditions for the Navier-Stokes equations, the vanishing of the total (viscous) energy dissipation (as the viscosity goes to zero) in a thin boundary layer (the so-called Prandtl-Von Karman layer) is equivalent to the validity of the vanishing viscosity limit. That is, the solutions of the Navier-Stokes equations converge, as the viscosity tends to zero, to the corresponding solution of the Euler equations, under the assumption of the existence of a strong solution of the Euler equations. This result is known as the Kato criterion and was established in [ 26 ] . A related criterion was then proved in [ 39 ] , which only required the vanishing of the part of the energy dissipation due to tangential derivatives of the velocity field (in a slightly thicker boundary layer than the one in [ 26 ] ). In [ 20 , 12 ] the regularity requirements on the