Regularity aspects of Leray–Hopf solutions to the 2D inhomogeneous Navier–Stokes system and applications to weak-strong uniqueness

Timothée Crin-Barat, Nicola De Nitti, Stefan Škondrić, Alessandro Violini · 2024

Read Regularity aspects of Leray–Hopf solutions to the 2D inhomogeneous Navier–Stokes system and applications to weak-strong uniqueness by Timothée Crin-Barat, Nicola De Nitti, Stefan Škondrić, Alessandro Violini (2024) 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-2412-13828.txt.

Introduction

Timothee Crin-Barat and co-authors sort smooth fixes to a fluid system in a 2024 paper. The system is the two dimensional Navier-Stokes flow with uneven density. Leray-Hopf fixes are the weak kind, built to hold for rough starts.

The team tell just which of those weak fixes turn strong after the start time. Their test rests on a strong energy rule and on traits of the pressure field. As a use, they prove that a weak fix and a strong fix with one start must match. The frame joins a string of fresh gains in the field in one clean view.

Catalog record

CREATORTimothée Crin-Barat, Nicola De Nitti, Stefan Škondrić, Alessandro Violini
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE98 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-13828

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Regularity aspects of Leray–Hopf solutions to the 2D inhomogeneous Navier–Stokes system and applications to weak-strong uniqueness Timothée Crin-Barat, Nicola De Nitti, Stefan Škondrić, Alessandro Violini Abstract . We characterize the Leray–Hopf solutions of the 2D inhomogeneous Navier–Stokes system that become strong for positive times. This characterization relies on the strong energy inequality and the regularity properties of the pressure. As an application, we establish a weak–strong uniqueness result and provide a unified framework for several recent advances in the field. 2020 Mathematics Subject Classification 76D03, 76D05. 1. Introduction 1.1. Presentation of the problem We consider the inhomogeneous incompressible Navier–Stokes system , which describes incompressible flows with nonconstant densities (see [ Lio96 , Chapter 1] ): (1.1) \displaystyle\begin{cases}\partial_{t}(\rho u)+\operatorname{div}(\rho u\otimes u)-\nu\Delta u+\nabla P=0,&(t,x)\in(0,\infty)\times\mathbb{R}^{d},\\ \partial_{t}\rho+\operatorname{div}(\rho u)=0,&(t,x)\in(0,\infty)\times\mathbb{R}^{d},\\ \operatorname{div}u=0,&(t,x)\in(0,\infty)\times\mathbb{R}^{d},\\ \rho(0,x)=\rho_{0}(x),&x\in\mathbb{R}^{d},\\ u(0,x)=u_{0}(x),&x\in\mathbb{R}^{d},\end{cases} where \rho\geq 0 is the density function , P is the pressure , u is the velocity field and \nu is the diffusivity of the fluid. System 1.1 satisfies the following scaling invariance: if (\rho,u,P) is a solution of 1.1 with initial data \left(\rho_{0},u_{0}\right) then, for all \lambda>0 , the rescaled triplet (1.2) \displaystyle\left(\rho_{\lambda},u_{\lambda},P_{\lambda}\right)\coloneqq\left(\rho\left(\lambda^{2}t,\lambda x\right),\lambda u\left(\lambda^{2}t,\lambda x\right),\lambda^{2}P\left(\lambda^{2}t,\lambda x\right)\right) is a solution of 1.1 with initial data (\rho_{0}(\lambda{{}\cdot{}}),\lambda u_{0}(\lambda{{}\cdot{}})) . Furthermore, provided that a solution (\rho,u,P) is sufficiently regular, the solution also conserves the energy: for every t\in(0,\infty) \displaystyle\frac{1}{2}\int_{\mathbb{R}^{d}}\rho(t)\left\lvert u(t,x)\right\rvert^{2}\,\mathrm{d}x+\nu\int_{0}^{t}\int_{\mathbb{R}^{2}}\left\lvert\nabla u(s,x)\right\rvert^{2}\,\mathrm{d}x\,\mathrm{d}s=\frac{1}{2}\int_{\mathbb{R}^{d}}\rho_{0}(x)\left\lvert u_{0}(x)\right\rvert^{2}\,\mathrm{d}x. Scale invariance suggests that the critical requirements for well-posedness are \rho_{0}\in L^{\infty}(\mathbb{R}^{d}) and u_{0}\in\dot{H}^{\frac{d}{2}-1}(\mathbb{R}^{d}) . In this work, we focus on the two-dimensional case ( d=2 ) with a strictly positive, bounded initial density and an L^{2} divergence-free initial velocity: (1.3) \displaystyle 0<c_{0}\leq\rho_{0}(x)\leq C_{0}<\infty\quad\text{for a.e.\penalty\ $x\in\mathbb{R}^{2}$}, \displaystyle u_{0}\in L^{2}_{\sigma}(\mathbb{R}^{2}). For d=2 , in [ Kaž74 ] , Kazhikhov established the existence Leray–Hopf weak solution for initial data 1.3 and, in [ Sim90 ] , Simon extended this result to the case \rho_{0}\geq 0 ( i.e. , allowing for possible vacuum in the initial density). 1 1 1 In [ Lio96 ] , Lions demonstrated that the density is a renormalized solution of the mass equation and considered the cases of density-dependent viscosity coefficients and unbounded densities as well. The uniqueness of Leray–Hopf weak solution of 1.1 remains an open question (while, in the constant density case, for d=2 , it was established by Ladyzhenskaya, Lions, and Prodi (see [ Lad59 , LP59 ] ). However, well-posedness results were proven assuming additional regularity on the initial data. In [ Dan03 , Dan04 ] , Danchin proved that, if u_{0}\in\dot{B}_{2,1}^{0}(\mathbb{R}^{2}) , then 1.1 is globally well-posed provided \rho_{0} is close to some positive constant in the homogeneous Besov space \dot{B}_{2,1}^{1}(\mathbb{R}^{2}) . In [ AG21 ] , Abidi and Gui proved a similar result for u_{0}\in\dot{B}_{2,1}^{0}\left(\mathbb{R}^{2}\right) and without the smallness condition but requiring \rho_{0}^{-1}-1\in\dot{B}_{2/\penalty\varepsilon,1}^{\varepsilon}(\mathbb{R}^{2}) for \varepsilon\in(0,1). For merely bounded initial density, bounded away from zero, well-posedness results were first obtained by Danchin and Mucha in [ DM13 ] and then improved by Paicu, Zhang, and Zhang in [ PZZ13 ] , where u_{0}\in H^{s}\left(\mathbb{R}^{2}\right) for some s>0 . Recently, in [ Hao+24 ] , Hao, Shao, Wei, and Zhang discussed the existence and uniqueness of solutions arising from initial data satisfying 1.3 . Additionally, in [ AMS24 ] , Adogbo, Mucha, and Szlenk constructed unique strong solutions for unbounded densities lying within a subspace of BMO. We recall that a Leray–Hopf weak solution with initial data satisfying 1.3 is a distributional solution of 1.1 that is energy admissible, meaning it satisfies the following energy inequality (see Definition 2.1 ): (1.4) \displaystyle\frac{1}{2}\int_{\mathbb{R}^{2}}\rho(t)|u(t)|^{2}\,\,\mathrm{d}x+\nu\int_{0}^{t}\int_{\mathbb{R}^{2}}|\nabla u(s)|^{2}\,\,\mathrm{d}x\,\,\mathrm{d}s\leq\frac{1}{2}\int_{\mathbb{R}^{2}}\rho_{0}|u_{0}|^{2}\,\,\mathrm{d}x\quad\text{for every $t\in(0,\infty)$}. Compared to the homogeneous model, several problems arise in the inhomogeneous case when analyzing Leray–Hopf weak solutions: • No control of the pressure is guaranteed for Leray–Hopf weak solutions to 1.1 : it is not even clear whether the pressure is a distribution. Unlike the homogeneous case, the regularity of the pressure cannot be recovered using Calderón–Zygmund-type estimates (as done in [ BV22 , Theorem 1.13] ). Indeed, the pressure P , at least for smooth triplet (\rho,u,P) satisfying 1.1 , solves (1.5) \displaystyle-\operatorname{div}\left(\frac{\nabla P}{\rho}\right)=\operatorname{div}\operatorname{div}(u\otimes u)-\operatorname{div}\left(\frac{\nu\Delta u}{\rho}\right), or, alternatively, the pressure can be recovered from the relation (1.6) \displaystyle-\Delta P=\operatorname{div}\operatorname{div}(\rho u\otimes u)+\operatorname{div}\left(\rho\partial_{t}u\right). In both ( 1.5 ) and ( 1.6 ), the interaction between P and \partial_{t}u or \Delta u makes it difficult to deduce any kind of regularity for P when \rho is merely bounded. • Secondly, it remains challenging to establish a strong version of the energy inequality ( 1.4 ). In the homogeneous case, applying Ladyzhenskaya’s inequality, 2 2 2 That is, Gagliardo–Nirenberg’s interpolation inequality \|D^{j}u\|_{L^{p}(\mathbb{R}^{d})}\leq C\|D^{m}u\|_{L^{r}(\mathbb{R}^{d})}^{\theta}\|u\|_{L^{q}(\mathbb{R}^{d})}^{1-\theta}, with d=2 , j=0 , m=1 , p=4 , q=2 , r=2 , and \theta=1/2 . we deduce u\in L^{4}((0,\infty)\times\mathbb{R}^{2}), and conclude that u satisfies the energy equality (see [ Gal00 , Theorem 4.1] ). The strategy is to test the weak formulation of the momentum equation using a spatial mollification of the velocity. However, in the inhomogeneous case, this approach cannot be implemented as it is unclear whether u , or any spatially mollified regularization of u , can be used as a test function in the weak formulation of 1.1 . 3 3 3 The main difficulty here stems from the lack of sufficient time regularity for the velocity field u . This issue arises because u is coupled with the density \rho , preventing effective use of the divergence-free condition. As a result, it remains an open question whether the same conclusion holds for non-constant densities. In particular, it is not even known if every Leray–Hopf solution (\rho,u) satisfies the strong energy inequality: for almost every s>0 and every t>s , (1.7) \displaystyle\frac{1}{2}\int_{\mathbb{R}^{2}}\rho(t)\left\lvert u(t)\right\rvert^{2}\,\,\mathrm{d}x+\nu\int_{s}^{t}\int_{\mathbb{R}^{2}}\left\lvert\nabla u(\tau)\right\rvert^{2}\,\,\mathrm{d}x\,\,\mathrm{d}\tau\leq\frac{1}{2}\int_{\mathbb{R}^{2}}\rho(s)\left\lvert u(s)\right\rvert^{2}\,\,\mathrm{d}x. • Due to the lack of a strong energy inequality, it is unclear whether there are any parabolic regularization effects. Given u_{0}\in H^{k}(\mathbb{R}^{2}) , with k\in\mathbb{N} , there exists a strong solution u of the homogeneous Navier–Stokes equations such that \displaystyle u\in L^{\infty}((0,\infty);H^{k}(\mathbb{R}^{2}))\cap L^{2}((0,\infty);\dot{H}^{k+1}(\mathbb{R}^{2})) (see [ RRS16 , Theorem 7.1] ). In addition, for a Leray–Hopf solution, we have that u(t)\in H^{1}(\mathbb{R}^{2}) for almost every t>0 . This implies the existence of a strong solution \tilde{u}_{t} with initial data u(t)\in H^{1}(\mathbb{R}^{2}) . By uniqueness, \widetilde{u}_{t} coincides with a t -time translation of u , leading to a gain in regularity for u for positive times: \displaystyle u\in L^{\infty}((t,\infty);H^{1}(\mathbb{R}^{2}))\cap L^{2}((t,\infty);\dot{H}^{2}(\mathbb{R}^{2})). By the freedom in the choice of t and iterating this procedure, we ultimately obtain \displaystyle u\in C^{\infty}((0,\infty)\times\mathbb{R}^{2}) (see [ RRS16 , Theorem 7.5] ). However, this result cannot be extended to Leray–Hopf solutions (\rho,u) of 1.1 with initial data satisfying ( 1.3 ). Although it is true that \displaystyle u(t)\in H^{1}(\mathbb{R}^{2})\quad\text{for almost every $t>0$,} and that there exists (owing to [ PZZ13 ] ) a strong solution (\widetilde{\rho}_{t},\widetilde{u}_{t}) with initial data (\rho(t),u(t)) such that (1.8) \displaystyle\widetilde{u}_{t}\in L^{\infty}((0,\infty);H^{1}(\mathbb{R}^{2}))\cap L^{2}((0,\infty);\dot{H}^{2}(\mathbb{R}^{2})), it is not clear how to conclude that (\widetilde{\rho}_{t},\widetilde{u}_{t}) is equal to a t -time translation of (\rho,u) . The main obstacle is that the weak-strong uniqueness argument (as presented, e.g. , in [ CŠV25 , Theorem 1.6] ) requires that (\rho,u) is a Leray–Hopf solution on [t,\infty)\times\mathbb{R}^{2} . In particular, we need the energy inequality for (\rho,u) in [t,\infty) and, as already discussed, this is not guaranteed for every Leray–Hopf weak solution. In our main result, Theorem 1.1 , we provide a connection

Read the complete text