Support growth of vorticity for bi-rotational Euler flows in high dimensions
In-Jee Jeong, Deokwoo Lim · 2025
Read Support growth of vorticity for bi-rotational Euler flows in high dimensions by In-Jee Jeong, Deokwoo Lim (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-2510-18335.txt.
Introduction
In-Jee Jeong and Deokwoo Lim study vorticity growth for spinning fluid flows in high dimensions. Vorticity measures local spinning in a fluid, and the Euler equations describe ideal flow without friction. Under bi-rotational symmetry without swirl, the equations shrink to one scalar equation in a single quadrant. For patch-type starting vorticity, the authors show the patch's support diameter grows without bound over time.
Catalog record
| CREATOR | In-Jee Jeong, Deokwoo Lim |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-18335
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Support growth of vorticity for bi-rotational Euler flows in high dimensions In-Jee Jeong, Deokwoo Lim Abstract . We study incompressible Euler equations in ℝ d \mathbb{R}^{d} with d ≥ 4 d\geq 4 under bi-rotational symmetry without swirl, which reduces the Euler equations to a scalar vorticity advection in the first quadrant. We show that patch type initial vorticities exhibit infinite growth of the support diameter. † † footnotetext: 2020 AMS Mathematics Subject Classification: 76B47, 35Q35 † † footnotetext: Key words: Vortex stretching; Bi-rotational symmetry; vorticity; Biot–Savart law 1. Introduction 1.1. Bi-rotational symmetry The incompressible Euler equations in \mathbb{R}^{d} are given by (1.1) \left\{\begin{aligned} \partial_{t}{\bf v}+{\bf v}\cdot\nabla{\bf v}+\nabla p&=0,\\ \nabla\cdot{\bf v}&=0,\end{aligned}\right. where {\bf v}=(v^{1},\cdots,v^{d}):[0,T)\times\mathbb{R}^{d}\to\mathbb{R}^{d} and p:[0,T)\times\mathbb{R}^{d}\to\mathbb{R} . In this paper, we take d\geq 4 and study ( 1.1 ) under the bi-rotational symmetry , using the bi-polar coordinates: we take two integers n,m\geq 1 where d=n+m+2\geq 4 , decompose \mathbb{R}^{d}=\mathbb{R}^{n+1}\times\mathbb{R}^{m+1} , and introduce two radii r,s\geq 0 by \begin{split}r^{2}=\sum_{i=1}^{n+1}x_{i}^{2},\qquad s^{2}=\sum_{i=n+2}^{n+m+2}x_{i}^{2}.\end{split} We define two unit vectors in \mathbb{R}^{d} by {\bf e}^{r}=\frac{1}{r}(x_{1},...,x_{n+1},0,...,0) and {\bf e}^{s}=\frac{1}{s}(0,...,0,x_{n+2},...,x_{n+m+2}) . We shall refer to bi-rotational symmetric solutions without swirl to be solutions of ( 1.1 ) of the form (1.2) {\bf v}(t,{\bf x})=u^{r}(t,r,s){\bf e}^{r}+u^{s}(t,r,s){\bf e}^{s},\qquad{\bf x}=(x_{1},...,x_{n+m+2}), where u^{r},u^{s} are scalar-valued. As long as the right-hand side of ( 1.2 ) satisfies some regularity and decay conditions at t=0 , we shall see that there is a unique corresponding local-in-time solution {\bf v}(t,{\bf x}) to ( 1.1 ), which verifies the ansatz ( 1.2 ). The term “bi-rotational” means that the velocity is invariant under all rotations of \mathbb{R}^{d} which either fixes the first n+1 coordinates or the last m+1 coordinates, and “without swirl” means that {\bf v}\in\mathrm{span}\{{\bf e}^{r},{\bf e}^{s}\} for every point in space-time. 1.2. Vorticity formulation of the bi-rotational solutions Recall that in \mathbb{R}^{d} the vorticity is a two-tensor \boldsymbol{\omega} defined from {\bf v} by (cf. [ 1 ] ) (1.3) \boldsymbol{\omega}=(\omega^{i,j})_{1\leq i,j\leq d},\quad\omega^{i,j}=\partial_{j}v^{i}-\partial_{i}v^{j}. The corresponding stream function is defined component-wise by (1.4) \Psi^{i,j}=\Delta^{-1}\omega^{i,j} where \Delta^{-1} is the inverse Laplacian operator defined in \mathbb{R}^{d} . Under the bi-rotational symmetry assumption on the velocity ( 1.2 ), there exist two scalar-valued functions of r and s satisfying (1.5) \omega^{i,j}({\bf x})=-\frac{x_{i}x_{j}}{rs}w(r,s),\qquad\Psi^{i,j}({\bf x})=-\frac{x_{i}x_{j}}{rs}\psi(r,s) for all 1\leq i\leq n+1,n+2\leq j\leq n+m+2 and {\bf x}\in\mathbb{R}^{d} . The proof is given in Section 2 . With an abuse of notation, we shall refer to w and \psi as the (scalar) vorticity and stream function, respectively. Furthermore, the Euler equations reduce to a scalar evolution equation for w : (1.6) \partial_{t}w+(u^{r}\partial_{r}+u^{s}\partial_{s})w=\bigg(n\frac{u^{r}}{r}+m\frac{u^{s}}{s}\bigg)w, or, (1.7) \partial_{t}\left(\frac{w}{r^{n}s^{m}}\right)+(u^{r}\partial_{r}+u^{s}\partial_{s})\left(\frac{w}{r^{n}s^{m}}\right)=0, where w:[0,T)\times\Pi\to\mathbb{R} with \Pi=\left\{(r,s)\,:\,r,s\geq 0\right\} . The bi-rotational velocity {\bf u}:=(u^{r},u^{s}) satisfies the divergence-free condition (defined in \mathbb{R}^{d} ) (1.8) \partial_{r}u^{r}+\frac{n}{r}u^{r}+\partial_{s}u^{s}+\frac{m}{s}u^{s}=0 as well as w=\partial_{r}u^{s}-\partial_{s}u^{r} . As such, using the stream function \psi which is related with w by (1.9) \left\{\begin{aligned} \bigg(\partial_{r}^{2}+\frac{n}{r}\partial_{r}-\frac{n}{r^{2}}+\partial_{s}^{2}+\frac{m}{s}\partial_{s}-\frac{m}{s^{2}}\bigg)\psi=w\quad&\mbox{in}\quad\Pi,\\ \psi=0\quad&\mbox{on}\quad\partial\Pi=\left\{(r,s)\,:\,r=0\mbox{ or }s=0\right\},\end{aligned}\right. we can recover {\bf u} from \psi from (1.10) u^{r}=-\frac{\partial_{s}(s^{m}\psi)}{s^{m}}=-\bigg(\frac{m}{s}\psi+\partial_{s}\psi\bigg),\quad u^{s}=\frac{\partial_{r}(r^{n}\psi)}{r^{n}}=\frac{n}{r}\psi+\partial_{r}\psi. The details of this computations are in the Section 2 . 1.3. Main result The form of the equation ( 1.7 ) shows that w/(r^{n}s^{m}) is simply being transported by {\bf u} . This allows us to consider the patch-type solutions, where w/(r^{n}s^{m}) is the indicator function of a (time-dependent) set in \Pi . We are interested in how the support size of w evolves for patch-type solutions of ( 1.7 ). Assuming that we are given a solution w\in L^{\infty}(0,T;L^{\infty}(\mathbb{R}^{d})) which is compactly supported in r and s for each t\in[0,T] , we define (1.11) \begin{split}R(t)&:=\sup\{r:(r,s)\in{\mathrm{supp}}\,w(t)\},\quad S(t):=\sup\{s:(r,s)\in{\mathrm{supp}}\,w(t)\},\quad L(t):=R(t)+S(t).\end{split} The following result shows local well-posedness for the patch-type data, when the support is initially separated from the axes. Theorem 1.1 . Take a bounded open set \Omega_{0}\subset\Pi which is separated from the boundary \partial\Pi of the domain; that is, \overline{\Omega_{0}}\cap\partial\Pi=\emptyset . Then, consider the patch-type initial datum (1.12) w_{0}(r,s)=r^{n}s^{m}\mathbf{1}_{\Omega_{0}}(r,s). There exist a time T>0 and a local unique solution of the Euler equations corresponding to ( 1.12 ), which also takes the form (1.13) w(t,r,s)=r^{n}s^{m}\mathbf{1}_{\Omega_{t}}(r,s),\quad t\in[0,T), where \Omega_{t}\in\Pi is the image of the set \Omega_{0} along the flow generated by {\bf u} in \Pi . Furthermore, the solution can be extended past T if and only if (1.14) \sup_{t\in[0,T)}L(t)<+\infty, where L(t) is defined in ( 1.11 ). Remark 1.1 . We note that for t<T , w\in L^{\infty}(\Pi) and this implies \boldsymbol{\omega}\in L^{\infty}(\mathbb{R}^{d}) . Moreover, \boldsymbol{\omega} is compactly supported in space and this guarantees that the corresponding velocity {\bf v} is log-Lipschitz in \mathbb{R}^{d} . In turn, {\bf u}=(u^{r},u^{s}) is log-Lipschitz in \Pi and w/(r^{n}s^{m}) is being transported by the flow map generated by {\bf u} in \Pi . Given \Omega_{0} satisfying the assumptions of Theorem 1.1 , we may define T_{max}=T_{max}(\Omega_{0})\in(0,+\infty] to be the maximal lifespan of the solution with initial datum ( 1.12 ). The following result shows that even when T_{max}=+\infty , we always have L(t)\to\infty along a sequence of time moments. Theorem 1.2 . Let w(t) be the local-in-time solution from Theorem 1.1 with the maximal time of existence T_{\text{max}}>0 . Then the support of w(t) satisfies (1.15) \sup_{t\in[0,T_{\text{max}})}L(t)=+\infty. Remark 1.2 . Using the time reversibility of the Euler equations, we can obtain the same conclusion when we solve the initial datum backwards in time. This corresponds to simply flipping the sign of the initial datum and solve forwards in time. In a similar vein, the conclusion also holds when \omega_{0}=\lambda r^{n}s^{m}\mathbf{1}_{\Omega_{0}}(r,s) for any constant \lambda\neq 0 . Unfortunately, the proof of Theorem 1.2 is largely based on a contradiction argument and it does not provide detailed information about the shape of the set \Omega_{t} as t\to T_{max} . For instance, one can ask what happens to the diameter, perimeter, curvature of the patch. Furthermore, we were not able to show that \|w(t,\cdot)\|_{L^{\infty}}\to+\infty as t\to\infty when T_{max}=+\infty . This is because along the flow trajectories, r^{n}s^{m} can remain bounded (or even decay) on \Omega_{t} . We leave these issues as open problems. 1.4. Discussion of the main results While it is known that the incompressible Euler equations in high dimensions can exhibit all kinds of dynamics ( [ 14 , 6 ] ), in general it is very difficult to analyze the solutions, as ( 1.1 ) is a system of d+1 scalar functions defined in \mathbb{R}^{d} . Assuming appropriate rotational symmetries on the high dimensional Euler solutions, one may reduce both the number of dependent and independent variables, while retaining some interesting dynamics. The most widely studied case is when the solution is assumed to be invariant under all rotations fixing an axis, and there have been several recent progress on high-dimensional axisymmetric Euler equations ( [ 3 , 13 , 8 , 10 , 11 ] ), concerning well-posedness, confinement, singularity formation, and so on. It seems that for higher dimensions, there is a better chance for finite-time singularity formation. On the other hand, dynamics under the bi-rotational symmetry ( 1.2 ) seems to have been less studied. The bi-rotational equation ( 1.7 ) appears in the works of Khesin–Yang [ 9 ] and Yang [ 15 ] (they use the terminology \mathbb{S}^{n}\times\mathbb{S}^{m} -symmetric solutions), where they noted that for any n,m\geq 1 , it is a special case of the so-called lake equations with the depth function d(r,s)=r^{n}s^{m} . The works [ 9 , 15 ]