Numerical study of transverse (in-)stability of solitary waves in the cubic-quintic nonlinear Schrödinger equation
Christian Klein, Christof Sparber · 2025
Read Numerical study of transverse (in-)stability of solitary waves in the cubic-quintic nonlinear Schrödinger equation by Christian Klein, Christof Sparber (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-2510-22735.txt.
Introduction
Solitary waves can lose stability when they bend across a channel. Christian Klein and Christof Sparber study that question in 2025. Their model is a nonlinear wave equation. It has cubic and quintic terms that compete. The domain is a line crossed with a ring.
The model is well posed in the global sense. It admits line shaped solitary waves. The pair test two kinds of disturbance. One is localized in space. The other repeats in a periodic way. Both tests point to a critical ring length. Above that length, the wave becomes unstable. Below it, the wave stays stable.
Catalog record
| CREATOR | Christian Klein, Christof Sparber |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-22735
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Numerical study of transverse (in-)stability of solitary waves in the cubic-quintic nonlinear Schrödinger equation Thanks: C.K. thanks for support by the ANR project ANR-17-EURE-0002 EIPHI and by the ANR project ISAAC-ANR-23-CE40-0015-01. C.S. is supported by the MPS Simons foundation through award no. 851720 Christian Klein, Christof Sparber Abstract . We study the nonlinear Schrödinger equation with a competing cubic-quintic power law nonlinearity on the waveguide domain ℝ x × 𝕋 L y \mathbb{R}_{x}\times\mathbb{T}_{L_{y}} . This model is globally well-posed and admits line solitary wave solutions, whose transverse (in-)stability is numerically investigated. We consider both spatially localized perturbations and periodic deformations of the line solitary wave and numerically confirm that there exists a critical torus length L y > 0 L_{y}>0 above which instability appears. 2000 Mathematics Subject Classification Primary: 35Q55. Secondary: 35C08. 1. Introduction In this work we consider the nonlinear Schrödinger equation (NLS) with competing cubic-quintic nonlinearity, i.e., (1.1) \left\{\begin{aligned} i\partial_{t}u+\Delta u&=-|u|^{2}u+|u|^{4}u,\\ u_{\mid t=0}&=u_{0}.\end{aligned}\right. This model appears in numerous applications, including nonlinear optics, superconductivity, and the description of Bose-Einstein condensates, see, e.g., [ 5 , 8 , 15 ] and the references therein. In contrast to the case of a single power-law nonlinearity, this model no longer admits a scaling symmetry, which makes the mathematical analysis more difficult. Indeed, ever since the seminal paper of Killip et al. [ 10 ] , this NLS has attracted considerable mathematical interest, cf. [ 2 , 7 , 13 , 16 , 20 ] . Equation ( 1.1 ) formally conserves (among others) the mass M(u)=\|u(t,\cdot)\|_{L^{2}}^{2} and the total energy E(u)=\frac{1}{2}\|\nabla u(t,\cdot)\|_{L^{2}}^{2}-\frac{1}{2}\|u(t,\cdot)\|_{L^{4}}^{4}+\frac{1}{3}\|u(t,\cdot)\|_{L^{6}}^{6}. Combined with Hölder’s inequality \|u\|_{L^{4}}^{4}\leqslant\|u\|_{L^{2}}\|u\|_{L^{6}}^{3}, these conservation laws imply that the cubic, focusing part cannot be an obstruction to global well-posedness for solutions u(t,\cdot) in the energy-space H^{1} . This is in sharp contrast to the case without the defocusing quintic nonlinearity, where finite-time blow-up of solutions is well-known in space dimensions two and higher, cf. [ 4 ] . 1.1. One- and two-dimensional ground states A particular class of solutions to ( 1.1 ) are solitary waves of the form u(t,{\bf x})=\phi_{\omega}({\bf x})e^{i\omega t},\quad\omega\in{\mathbb{R}}, where \phi_{\omega} is a localized (i.e., vanishing at infinity) solution to the stationary equation (1.2) -\Delta\phi_{\omega}+\omega\phi_{\omega}-|\phi_{\omega}|^{2}\phi_{\omega}+|\phi_{\omega}|^{4}\phi_{\omega}=0. By deriving the associated Pohozaev-Nehari identities, one can prove that a necessary condition for the existence of nontrivial profiles 0\not=\phi_{\omega}\in H^{1} is given by \omega\in(0,\frac{3}{16}) , see [ 2 ] . In one space dimension, the profile \phi_{\omega} can be explicitly computed in the form (1.3) \phi_{\omega}(x)=\left(\frac{1}{4\omega}+\sqrt{\frac{1}{16\omega^{2}}-\frac{1}{3\omega}}\cosh\big(2\sqrt{\omega}x\big)\right)^{-1/2},\quad x\in{\mathbb{R}}. In Fig. 1 we show the mass M(\phi_{\omega}) and the energy E(\phi_{\omega}) of this explicit family of 1D solitary waves as a function of \omega\in(0,\frac{3}{16}) . It can be seen that both M(\phi_{\omega})\geqslant 0 and E(\phi_{\omega})\leqslant 0 vanish as \omega\to 0 , while they both diverge as \omega\to\frac{3}{16} . Moreover, both are monotonic w.r.t. \omega . Figure 1. The mass (left) and the energy (right) of the 1D solitary wave ( 1.3 ) as a function of \omega . Let us briefly recall the definition of orbital stability of solitary waves. The need of the inclusion of the parameters \theta\in{\mathbb{R}} and {\bf x}_{0}\in{\mathbb{R}}^{d} is thereby due to the gauge and translation invariance of ( 1.1 ). Definition 1.1 . We say that the standing wave u(t,{\bf x})=e^{i\omega t}\phi_{\omega}({\bf x}) is orbitally stable, if for all \varepsilon>0 , there exists \delta>0 such that if u_{0}\in H^{1} satisfies \|u_{0}-\phi_{\omega}\|_{H^{1}}\leqslant\delta, then the solution to ( 1.1 ) with u_{\mid t=0}=u_{0} satisfies \sup_{t\in{\mathbb{R}}}\inf_{{\theta\in{\mathbb{R}}}\atop{{\bf x}_{0}\in{\mathbb{R}}^{d}}}\left\|u(t,\cdot)-e^{i\theta}\phi_{\omega}(\cdot-{\bf x}_{0})\right\|_{H^{1}}\leqslant\varepsilon. Otherwise, the standing wave is said to be unstable. In view of Fig. 1 , we have that \frac{\partial}{\partial\omega}M(\phi_{\omega})>0 for the family of profiles ( 1.3 ). The well-known Vakhitov and Kolokolov stability criterion [ 22 ] therefore indicates orbital stability of these 1D solitary waves. A rigorous proof of this fact was given in [ 20 ] , and we also refer to [ 9 , 23 ] for a more general mathematical theory concerning the (in-)stability of standing waves. In higher dimensions, no explicit formula for profiles \phi_{\omega}({\bf x}) with {\bf x}\in{\mathbb{R}}^{d} , is known. However it can be proved that for any given \omega\in(0,\frac{3}{16}) , there exists a unique (up to gauge invariance) solution \phi_{\omega}({\bf x})=Q_{\omega}(|{\bf x}|)\geqslant 0 , i.e., a radially symmetric, non-negative profile Q_{\omega} , which, together with its partial derivatives up to order two, decays exponentially to zero as r\equiv|{\bf x}|\to\infty , see, e.g., [ 2 ] for more details. In 1D we clearly have (by uniqueness) that Q_{\omega} is identical to the profile \phi_{\omega} given in ( 1.3 ). In 2D these (cubic-quintic) ground state solutions Q_{\omega} have been numerically constructed in [ 3 ] . For the convenience of the reader, we shall briefly recall some of its findings: In Fig. 2 we show a plot of the 2D profiles Q_{\omega} and their L^{\infty} -norms for several values of \omega\in(0,\frac{3}{16}) . Figure 2. Left: Numerically constructed 2D ground state solutions to ( 1.1 ) for several values of \omega . Right: The L^{\infty} -norm of these states as a function of \omega . In Fig. 3 we show the (numerically computed) mass and the energy of these 2D solitary wave solutions as functions of \omega . Both quantities are seen to remain monotonic with \frac{\partial}{\partial\omega}M(Q_{\omega})>0 , just as in the 1D case above. Thus, one again expects orbital stability of the whole family of 2D ground states [ 3 , 13 ] . However, no rigorous proof of this is available to date (see [ 3 ] for a numerical study of the long-time behavior of perturbed ground states in 2D and 3D). Figure 3. M(Q_{\omega}) and E(Q_{\omega}) as functions of \omega in 2D. 1.2. Transverse (in-)stability of line solitary waves In this work, we are interested in an in-between situation, namely the transverse (in-)stability of the profiles ( 1.3 ). Clearly, the solution \phi_{\omega}=\phi_{\omega}(x) with x\in{\mathbb{R}} can be seen as a particular kind of 2D solution to ( 1.1 ) which is constant in the transverse variable y , a so-called line soliton . For {\bf x}=(x,y)\in{\mathbb{R}}^{2} such (partially constant) solutions have infinite mass and energy, which is why we shall instead consider ( 1.1 ) on a cylindrical domain with (x,y)\in\mathbb{R}_{x}\times\mathbb{T}_{L_{y}} . Here \mathbb{T}_{L_{y}}={\mathbb{R}}/(2\pi L_{y}{\mathbb{Z}}) denotes the torus of length L_{y}>0 . This setting is physically meaningful as it describes propagation within waveguide structures. The well-posedness theory for NLS on such waveguide manifolds is by now well understood (albeit not yet as complete as for the usual Euclidian setting), see, e.g., [ 7 , 14 , 21 , 26 ] and the references therein. In our setting we are consequently guaranteed the existence of global in-time solutions u\in C({\mathbb{R}};H^{1}(\mathbb{R}_{x}\times\mathbb{T}_{L_{y}})) to ( 1.1 ), depending continuously on the initial data u_{0} . The definition of transverse (orbital) stability of line solitons u(t,x)=\phi_{\omega}(x)e^{i\omega t},\quad\omega\in(0,\tfrac{3}{16}), under the evolution of a 2D nonlinear flow is thereby the same as given in Definition 1.1 , except that the parameter {\bf x}_{0}\in{\mathbb{R}}^{d} has to be replaced by x_{0}\in{\mathbb{R}} . Rousset and N. Tzvetkov studied these type of problems in a series of seminal works [ 17 , 18 , 19 ] in which they developed a rather general mathematical framework for the transverse (in-)stability of solitary waves to dispersive equations in 2D. More recent studies of