A Besov-based integration-by-parts method for the incompressible Navier-Stokes equations
Xinyu Cheng, Zhaonan Luo, Sheng Wang · 2025
Read A Besov-based integration-by-parts method for the incompressible Navier-Stokes equations by Xinyu Cheng, Zhaonan Luo, Sheng Wang (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-2509-23192.txt.
Introduction
Fluid equations with rough solutions need error bounds that respect that roughness. Xinyu Cheng, Zhaonan Luo, and Sheng Wang analyze a Navier-Stokes scheme in Besov spaces. Besov spaces grade functions by finer smoothness scales than classic Sobolev spaces. The team study a semi-implicit time-stepping scheme for incompressible flow. They derive precise error estimates in two Besov settings.
One case needs a careful integration-by-parts trick. That trick shifts derivatives onto test functions to tame the nonlinear advection term. The bounds tie convergence to the critical regularity of the true solution. The authors argue Besov tools give sharper, more local error control.
Catalog record
| CREATOR | Xinyu Cheng, Zhaonan Luo, Sheng Wang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-23192
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A Besov-based integration-by-parts method for the incompressible Navier-Stokes equations Xinyu Cheng, Zhaonan Luo, Sheng Wang Abstract . This note introduces a novel numerical analysis framework for the incompressible Navier-Stokes equations based on Besov spaces. The key contribution of this note is to establish the stability and convergence of a semi-implicit time-stepping scheme by deriving precise error estimates in the B ∞ , 1 0 B^{0}_{\infty,1} and B ∞ , 2 0 B^{0}_{\infty,2} spaces. Another contribution of our analysis is the detailed treatment of the B ∞ , 2 0 B^{0}_{\infty,2} case, where a crucial integration-by-parts technique is employed to adeptly handle the nonlinear advection term. This technique allows for a refined estimate that effectively transfers derivatives onto the test functions, mitigating the inherent analytical challenges posed by the low regularity of these spaces. Our results provide sharper, more localized error bounds than in classical Sobolev spaces, directly linking the scheme’s convergence to the critical regularity of the continuous solution. This work underscores the advantage of Besov spaces for the numerical analysis of nonlinear fluid PDEs. 1. Introduction The numerical analysis of the incompressible fluid equations remains a topic of profound and enduring interest, standing at the crossroads of theoretical mathematics (partial differential equations and analysis) and computational fluid dynamics (engineering). While the (global) well-posedness theory for these equations is famously challenging, the development of robust and efficient numerical schemes is equally critical for predictive simulations across science and engineering. A vast body of literature is devoted to constructing discretization methods, with a predominant focus on error estimates formulated within the classical Sobolev spaces, such as L^{2}(\Omega) and H^{s}(\Omega) . These frameworks, though powerful, can sometimes obscure finer-grained information about the solution, particularly concerning its pointwise behavior or spaces closely related to the well-posedness theory of the equations themselves. In this paper, we introduce a novel numerical analysis framework grounded in the theory of Besov spaces. Specifically speaking, we will perform a rigorous error analysis for a semi-implicit time-stepping scheme applied to the Navier-Stokes equations, deriving error bounds in the B^{0}_{\infty,1} and B^{0}_{\infty,2} norms. Here the incompressible Navier-Stokes system (NS) we consider are described as below: (1.1) \begin{cases}&\partial_{t}u+u\cdot\nabla u+\nabla p=\nu\Delta u,\qquad\nabla\cdot u=0;\\ &u|_{t=0}=u_{0}.\end{cases} The solution u:\mathbb{R}^{+}\times\Omega\to\mathbb{R}^{d} is the velocity field and p:\mathbb{R}^{+}\times\Omega\to\mathbb{R} is the pressure. Moreover, the divergence-free condition \nabla\cdot u=0 is to guarantee the incompressiblity. Since we are interested in the analysis framework, for the sake of implicit we shall only consider the periodic boundary condition: \Omega=\mathbb{T}^{2} . And the semi-implicit scheme we consider is the following: (1.2) \begin{cases}&\frac{u^{n+1}-u^{n}}{\tau}+(u^{n}\cdot\nabla)u^{n+1}+\nabla p^{n}=\nu\Delta u^{n+1},\\ &\mathop{\rm div}u^{n+1}=0,\end{cases} where \tau is the time step. Moreover in ( 1.2 ) the viscosity term \nu\Delta u^{n+1} is unknown and the nonlinear term u^{n}\cdot\nabla u^{n+1} is indeed linear with respect to u^{n+1} at n+1 -th time step. In addition we can further derive that \mathop{\rm div}u^{n}=0 for all n by induction. This choice of functional setting is motivated by several compelling factors. Firstly, Besov spaces like B^{0}_{\infty,1} are known to be critical for the existence and uniqueness theory of Navier-Stokes and Euler equations; indeed the global well-posedness of the Cauchy problem of the Euler equations in the borderline Besov space B^{\frac{3}{p}+1}_{p,1}(\mathbb{R}^{3}) for p\in[1,\infty] is known to be open, cf. Bourgain-Li [ 6 , 7 ] . These spaces provide a natural and less restrictive environment for analyzing solutions compared to Sobolev spaces, particularly for problems involving rough initial data or intricate vortex structures. Secondly, an error estimate in B^{0}_{\infty,\cdot} offers a stronger, more localized description over the pointwise error of the numerical solution than an L^{2} -estimate. It is worth mentioning here in a recent work [ 12 ] , a numerical method of low regularity assumption was developed to study the surface quasi-geostrophic equations in a B^{0}_{\infty,1} set-up, which motivates us to develop a more general B^{0}_{\infty,1} framework for impressible fluid equations such as Navier-Stokes and Euler equations in this presenting note. Moreover, in another recent work [ 11 ] we observe that a simple integration-by-parts can help to lower the regularity assumption, which motivates us to study the B^{0}_{\infty,2} numerical space. Our work demonstrates that the proposed semi-implicit scheme is stable and convergent within this refined functional framework. The analysis reveals how the regularity of the exact solution, measured in these Besov norms, directly dictates the convergence rates of the numerical method. By bridging the gap between the analytical functional setting of the continuous equations and the discrete numerical analysis, this approach not only provides stronger theoretical guarantees but also offers deeper insight into the structure of the error. We anticipate that this Besov-based framework will open new avenues for the analysis and development of numerical methods for a broader class of nonlinear fluid PDEs where such spaces are intrinsic to the well-posedness theory. We now give a brief review on the literature of interest. First of all, we refer the well-posedness theory of the Euler and Navier-Stokes equations to [ 2 , 50 , 4 , 35 ] and the references therein. We also refer the readers to [ 13 , 34 ] for more discusssion in dfferent models and function spaces. Since we are also interested in developing the vanishing-viscosity-limit structure-preserving schemes, we list here several related results for different fluid models in different function spaces: [ 12 , 22 , 30 , 41 ] . The ill-posednes of the different fluid systems have been investigated, cf. [ 6 , 7 , 8 ] . The non-uniqueness of different fluid equations have been obtained in [ 9 , 10 , 16 ] and etc. On the other hand, there are many numerical methods to study the incompressible fluid models including Euler and Navier-Stokes equations. For example, the backward Euler differentiation method cf. [ 24 , 47 ] , finite element methods cf. [ 27 , 28 , 29 , 26 , 46 ] , finite difference methods cf. [ 14 , 17 , 18 , 48 ] , spectral methods cf. [ 21 , 11 ] , the Lagrange–Galerkin method cf. [ 25 , 43 ] , the projection method cf. [ 15 , 31 , 17 , 18 ] , exponential low regularity methods cf. [ 37 , 42 , 3 , 49 , 12 ] and neural network based methods cf. [ 23 , 32 , 33 ] . We also refer the readers to [ 40 , 44 , 45 ] and the references therein for a more detailed discussion. Now in pratice to solve u^{n+1} from ( 1.2 ), we recall the Leray projection \mathbb{P} , namely the L^{2} -orthogonal projection onto the divergence-free subspace: for any v\in L^{2} we have \mathbb{P}v=v-\nabla q , where q\in H^{1} solves the following Poisson equation under periodic boundary conditions: \Delta q=\nabla\cdot v. Then we apply the Leray projection \mathbb{P} to derive that (1.3) \frac{u^{n+1}-u^{n}}{\tau}+\mathbb{P}(u^{n}\cdot\nabla u^{n+1})=\nu\Delta u^{n+1}. In fact we will solve the NS (Euler) equations from the first order semi-implicit scheme ( 1.3 ). We shall show our scheme ( 1.3 ) preserves regularity stability; moreover the error will be shown with numerical evidence and rigorous analysis. Our main results state below. To start with, we first present the stability result of the semi-implicit scheme ( 1.2 ) or ( 1.3 ). Theorem