Lawson cones and the Allen-Cahn equation
Oscar Agudelo, Matteo Rizzi · 2025
Read Lawson cones and the Allen-Cahn equation by Oscar Agudelo, Matteo Rizzi (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-2501-15812.txt.
Introduction
Oscar Agudelo and Matteo Rizzi study special minimal surfaces that approach a Lawson cone at large distances. Their 2025 paper first checks how stable those surfaces are and whether nearby surfaces can bend them. Minimal surfaces are shapes that locally take the least area, like soap films. The authors then use these surfaces to build solutions of a known equation whose zero set has several separate parts and an endless index of instability.
Catalog record
| CREATOR | Oscar Agudelo, Matteo Rizzi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-15812
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Lawson cones and the Allen-Cahn equation Oscar Agudelo, Matteo Rizzi Abstract . In this paper we discuss nondegeneracy and stability properties of some special minimal hypersurfaces which are asymptotic to a given Lawson cone C m , n C_{m,n} , for m , n ≥ 2 m,\,n\geq 2 . Then we use such hypersurfaces to construct solutions to the Allen-Cahn equation − Δ u = u − u 3 -\Delta u=u-u^{3} in ℝ N + 1 \mathbb{R}^{N+1} , N + 1 ≥ 8 N+1\geq 8 , whose zero level set has exactly k ≥ 2 k\geq 2 connected components and with infinite Morse index. Keywords. Allen-Cahn equation, Lawson cones, Minimal Surfaces, Jacobi-Toda system. 1. Introduction In this work we discussed part of the recent developments related to entire and bounded solutions to the Allen-Cahn equation (1.1) \Delta u-F^{\prime}(u)=0\quad\hbox{in}\quad\mathbb{R}^{N+1}, F(s)=\frac{1}{4}(1-s^{2})^{2} for s\in\mathbb{R} is the best example of a symmetric double well potential . Equation ( 1.1 ) appears in the modeling of phase transitions phenonema (see [ 7 ] ). Roughly speaking, the functions u_{+}=1 and u_{-}=-1 represent two different phases of a material placed in a container. When the two phases co-exist (not mixing involved), they are separated by an interface. The equilibrium configuration is the smooth function u , which describes the phase change, and is expected to be a solution of ( 1.1 ). Observe that u_{\pm} are trivial solutions of ( 1.1 ) corresponding to stable phases of the said material. Theoretical developments related to the equation ( 1.1 ) are known to be strongly connected with the Theory of Minimal Surfaces . This statement can be briefly explained as follows. First, we remark that in dimension one, the function (1.2) w(z)=\tanh\left(\frac{z}{\sqrt{2}}\right),\quad z\in\mathbb{R}. is the unique solution (up tu translation of the argument) to the problem (1.3) w^{\prime\prime}+w(1-w^{2})=0,\quad\hbox{in }\mathbb{R},\quad w(\pm\infty)=\pm 1,\quad w^{\prime}>0. It is verified that \small{w(z)=\left\{\begin{array}[]{ccc}1-2e^{-\sqrt{2}\,z}+O\left(e^{-2\sqrt{2}\,z}\right),\hskip 9.24994ptz>>0\\ -1+2e^{-\sqrt{2}\,z}+O\left(e^{-2\sqrt{2}\,z}\right),\hskip 9.24994ptz<<0.\\ \end{array}\right.} Next, the equation ( 1.1 ) appears as the limit of its singularly perturbed version, namely (1.4) \epsilon^{2}\Delta v-F^{\prime}(v)=0\quad\hbox{in }\Omega,\qquad\ \frac{\partial v}{\partial\nu}=0\quad\hbox{on }\partial\Omega,\qquad\epsilon\to 0, where \Omega\subset\mathbb{R}^{N+1} is a smooth bounded domain. As discussed in [ 19 ] , given a solution v of ( 1.4 ), the rescaling u_{\epsilon}(x)=v(\epsilon x) yields that \Delta u_{\epsilon}-F^{\prime}(u_{\epsilon})=0\quad\hbox{in}\quad\Omega_{\epsilon}:=\epsilon^{-1}\Omega. Since, \epsilon^{-1}\Omega\to\mathbb{R}^{N} , from standard Schauder estimates, in the limit we obtain a smooth bounded function u that solves ( 1.1 ). On the other hand, regarding the solutions of the equation ( 1.4 ), they are exactly the critical points in H^{1}(\Omega) of the energy J_{\epsilon}(v)=\int_{\Omega}\frac{\epsilon}{2}\,|\nabla v|^{2}+\frac{1}{\epsilon}\,F(v). Observe that v=\pm 1 are global minimizers of J_{\epsilon} . The capture of configurations with coexisting phases is based on the intuition that such configurations v_{\epsilon} should resemble an \epsilon- regularization of the function v^{*}=\mathds{1}_{\Lambda}-\mathds{1}_{\Omega\setminus\Lambda}, where \Lambda\subset\Omega . The characteristic functions \mathds{1}_{\Lambda} and \mathds{1}_{\Omega\setminus\Lambda} of \Lambda and \Omega\setminus\Lambda respectively might be thought as constant phases of a material. Observe also that v^{*} minimizes the second term in J_{\epsilon} , but v^{*}\notin H^{1}(\Omega) (see [ 19 ] ). Let M denote the boundary of \Lambda relatively to \overline{\Omega} . The ideas in [ 19 ] are based on the intuition that when M is a smooth orientable hypersurface, the transition from the phase \mathds{1}_{\Lambda} to the phase \mathds{1}_{\Omega\setminus\Lambda} takes place along the normal direction of M with a approximate one dimensional profile, i.e., v_{\epsilon}(x)\approx w\left(\frac{z}{\epsilon}\right), where z corresponds to the normal direction to M and w is described in ( 1.2 ) and ( 1.3 ). The gradient term in J_{\epsilon}(v_{\epsilon}) suggests that M should be critical for the area. To make this idea slightly more precise, we take \epsilon\to 0^{+} , so that J_{\epsilon}(v_{\epsilon})\approx{\rm Area}(M)\int_{\mathbb{R}}\left(\frac{1}{2}|w^{\prime}|^{2}+F(w)\right)dz. This intuition lead to Modica’s result in [ 19 ] , which states that for a family of local minimizers \{v_{\epsilon}\}_{\epsilon} of J_{\epsilon} with bounded energy, v_{\epsilon}\to v^{*}\quad\hbox{in}\quad L^{1}_{loc}, where M=\partial\Lambda has minimal perimeter. We remark that in case M is a smooth hypersurface, this means that M is a minimal hypersurface. Observe also that M is the approximate nodal set of v_{\epsilon} for \epsilon>0 small. Kohn and Sternberg in [ 17 ] constructed a family \{v_{\varepsilon}\}_{\varepsilon}\in(0,\varepsilon_{0}) of solutions of ( 1.4 ) on a bounded domain \Omega\subset\mathbb{R}^{2} whose zero level set is approaching some given segment as \varepsilon\to 0 . A similar result on compact manifolds was proved by Pacard and Ritoré [ 21 ] on more general manifolds. Modica’s developments and many other related developments during the 70th’s, lead De Giorgi to formulate its remarkable conjecture in 1978. De Giorgi conjecture (1978): Let u\in L^{\infty}(\mathbb{R}^{N}) be an entire solution of the equation. Assume also that u is monotone in one direction and that 2\leq N+1\leq 8 . Then, after rotation and translation u(x)=w(x_{1}) , where is determined by ( 1.3 ). Some of the key developments in proving or disproving the De Giorgi’s conjecture are the following. This is motivated by the fact that entire minimal graphs over \mathbb{R}^{N+1} are affine if N\leq 7 ( Bernstein conjecture ). The conjecture is known to be true in dimension N+1=2,\,3 (see [ 15 , 8 ] ). In dimension 4\leq N+1\leq 8 only partial results are available. For instance Savin [ 22 ] proved that in those dimensions the conjecture is true under the further assumption that u(x_{1},x^{\prime})\to\pm 1 as x_{1}\to\pm\infty , for any x^{\prime}\in\mathbb{R}^{N} . Moreover, it is known that the conjecture is sharp about the dimension. In fact in [ 13 ] the authors constructed a family of monotone solutions in \mathbb{R}^{9} which are not one-dimensional. At the core of the De Giorgi’s conjecture is the hypothesis of monotonicity of the solutions, which implies their stability. To be more precise, let u\in L^{\infty}(\mathbb{R}^{N}) be a solution of ( 1.1 ) and consider the linearized operator of ( 1.1 ) around u , namely L\psi:=\Delta\psi+(1-3u^{2})\psi\quad\hbox{in }\mathbb{R}^{N},\qquad\psi\in L^{\infty}(\mathbb{R}^{N}). The stability of u is equivalent to saying that the operator L satisfies the maximum principle or, in turn it is equivalent to say that the associated quadratic form B(\psi,\psi):=\int_{\mathbb{R}^{N}}|\nabla\psi|^{2}-(1-3u^{2})\psi^{2}. is positive definite in the space C_{c}^{\infty}(\mathbb{R}^{N}) . The Morse Index of u , m(u) , is the maximal dimension of a vector space of smooth functions \psi with compact support for which B(\psi,\psi)<0. A solution u\in L^{\infty}(\mathbb{R}^{N}) of ( 1.1 ) is unstable if m(u)>0 . Regarding unstable solutions the following are some of the recent and relevant examples and results. Dang, Fife, Peletier in [ 12 ] exhibited saddle solutions in dimension two. Their result was later generalized to higher dimensions by Cabre and Terra, first in 2010 by showing unstable saddle solutions and then in 2011 by showing stable saddle solutions (see [ 10 , 11 ] ). In 2011, del Pino, Kowalczyk, Wei in [ 14 ] constructed solutions to ( 1.1 ) in three dimensions. The nodal set of such solutions is close to a large dilation of a complete embedded non-degenerate minimal surface with finite total curvature. These solutions also have finite Morse Index. As a follow up for the three dimensional setting, in 2012 Agudelo, del Pino and Wei constructed two types of solutions to ( 1.1 ), one with Morse Index one and disconnected nodal set asymptotically characterized by Liouville’s equation. The other one with Morse index going to infinity and nodal