Singular minimal surfaces with constant curvature

Rafael López · 2025

Read Singular minimal surfaces with constant curvature by Rafael López (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-2507-13668.txt.

Introduction

Rafael Lopez classifies a special family of surfaces that combine least-area behaviour with a weighted area rule, in a 2025 paper. He asks which such surfaces can keep a fixed Gauss curvature, the measure of how a surface bends in two directions at once. The answer turns out to be short and strict. Only planes, spheres, and cylinders qualify. Lopez also sorts all surfaces in the family holding one principal curvature fixed, and all holding mean curvature fixed.

Catalog record

CREATORRafael López
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-13668

Browse the full library · About the Tower of Babel

Full text

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

Singular minimal surfaces with constant curvature Rafael López Abstract . We prove that singular minimal surfaces with constant Gauss curvature are planes, spheres and cylindrical surfaces. We also classify all singular minimal surfaces with a constant principal curvature and singular minimal surfaces with constant mean curvature. 1991 Mathematics Subject Classification 53A04, 53B25 1. Introduction and statement of the results Let \vec{a}\in\mathbb{R}^{3} be a unit vector in Euclidean space \mathbb{R}^{3} and \alpha\in\mathbb{R} . Let \Sigma be an oriented surface and \Phi\colon\Sigma\to\mathbb{R}^{3} be an isometric immersion of \Sigma in \mathbb{R}^{3} . The surface \Sigma is said to be an \alpha -singular minimal surface if its mean curvature H satisfies H=\alpha\frac{\langle N,\vec{a}\rangle}{\langle\Phi,\vec{a}\rangle}, (1.1) where N is the unit normal vector of \Sigma . We are implicitly assuming that \Phi(\Sigma) is included in one of the two halfspaces determined by the vector plane \Pi=\{p\in\mathbb{R}^{3}\colon\langle p,\vec{a}\rangle=0\} . Singular minimal surfaces are critical points of the energy functional \Sigma\mapsto\int_{\Sigma}\langle\Phi(p),\vec{a}\rangle^{\alpha}\,d\Sigma, where d\Sigma is the area element of \Sigma . If \alpha=0 , this energy is simply the area functional and the critical points are minimal surfaces. From now, we discard the case \alpha=0 . The case \alpha=1 extends in dimension two the notion of catenary: see [ 3 , 8 , 15 ] . The study of the singular minimal surface equation can be viewed in [ 2 , 4 , 5 , 12 , 13 ] . For geometric aspects of singular minimal surfaces, see [ 7 , 9 , 11 , 14 ] . Among the examples of \alpha -singular minimal surfaces, we focus on isoparametric surfaces, that is, planes, spheres and circular cylinders. (1) Planes parallel to \vec{a} are \alpha -singular minimal surfaces for all \alpha\in\mathbb{R} . (2) Spheres are \alpha -singular minimal surfaces only if \alpha=-2 . In such a case, the center of the sphere lies in the vector plane \Pi . These surfaces also represent minimal surfaces in hyperbolic space \mathbb{H}^{3} when \mathbb{H}^{3} is viewed in the upper halfspace model \mathbb{R}^{3}_{+}=\{p\in\mathbb{R}^{3}\colon\langle p,\vec{a}\rangle>0\} . (3) Circular cylinders are \alpha -singular minimal surfaces only if \alpha=-1 . In such a case, the rotational axis lies in the vector plane \Pi . Other examples of \alpha -singular minimal surfaces can be found in the family of cylindrical surfaces, that is, ruled surfaces whose rulings are parallel straight-lines. If the cylindrical surface is parametrized by \Psi(s,t)=\gamma(s)+t\vec{v} , |\vec{v}|=1 , and if \Sigma satisfies ( 1.1 ), then either \Sigma is a plane parallel to \vec{a} or \langle\vec{v},\vec{a}\rangle=0 . This implies that the generating curve \gamma lies contained in a plane parallel to \vec{a} . In such a case, Eq. ( 1.1 ) writes in terms of the curvature \kappa of \gamma as \kappa=\alpha\frac{\langle{\bf n},\vec{a}\rangle}{\langle\gamma,\vec{a}\rangle}, (1.2) where {\bf n} is the unit normal of \gamma . These surfaces have zero Gauss curvature. Cylindrical singular minimal surfaces have been classified in [ 9 , 11 ] . If \alpha=-1 , cylindrical singular minimal surfaces are circular cylinders. In this paper, we will refer these surfaces simply by cylindrical singular minimal surfaces. Planes parallel to \vec{a} , spheres centered at \Pi and the cylindrical singular minimal surfaces have constant Gauss curvature K . The purpose of this paper is to prove the converse and characterize these surfaces are the only singular minimal surfaces with constant Gauss curvature. Theorem 1.1 . Planes parallel to \vec{a} , spheres centered at \Pi and the cylindrical singular minimal surfaces are the only \alpha -singular minimal surfaces with constant Gauss curvature. This result will proved in Sect. 3 . We also study those singular minimal surfaces with a constant principal curvature. In Sect. 4 , we prove the following classification. Theorem 1.2 . Planes parallel to \vec{a} , spheres centered at \Pi and cylindrical singular minimal surfaces are the only singular minimal surfaces with a constant principal curvature. Finally we consider singular minimal surfaces with constant mean curvature. Theorem 1.3 . Planes parallel to \vec{a} and spheres centered at \Pi are the only singular minimal surfaces with constant mean curvature. 2. Preliminaries In this section we describe parametrizations of a surface by lines of curvature. This will be done in open sets of a surface free of umbilic points. Let \Sigma be an orientable surface immersed in \mathbb{R}^{3} . Denote by \nabla and \overline{\nabla} be the Levi-Civita connections of \Sigma and \mathbb{R}^{3} , respectively. Denote by N the unit normal vector field of \Sigma . Let \mathfrak{X}(\Sigma) be the space of tangent vector fields of \Sigma . The Gauss and Weingarten formulas are, respectively, \begin{split}\overline{\nabla}_{X}Y&=\nabla_{X}Y+h(X,Y),\\ \overline{\nabla}_{X}N&=-SX\end{split} (2.1) for all X,Y\in\mathfrak{X}(\Sigma) , where h\colon T\Sigma\times T\Sigma\to T^{\perp}\Sigma and S\colon T\Sigma\to T\Sigma are the second fundamental form and the shape operator, respectively. The Gauss equation and Codazzi equations are \begin{split}\langle R(X,Y)W,Z\rangle&=\langle h(Y,W),h(X,Z)\rangle-\langle h(X,W),h(Y,Z)\rangle\\ (\nabla_{X}S)Y&=(\nabla_{Y}S)X,\end{split} where R is the curvature tensor of \nabla . The mean curvature H and the Gauss curvature K of \Sigma are defined by H=\kappa_{1}+\kappa_{2}\quad K=\kappa_{1}\kappa_{2}, where \kappa_{1} and \kappa_{2} are the principal curvatures of \Sigma . In an open set \Omega\subset\Sigma of non-umbilic points, consider a local orthonormal frame \{e_{1},e_{2}\} given by coordinates of lines of curvature. The expression of the shape operator S is Se_{1}=\kappa_{1}e_{1},\qquad Se_{2}=\kappa_{2}e_{2}. (2.2) Since \{e_{1},e_{2}\} is an orthonormal frame, if X\in\mathfrak{X}(\Sigma) then \nabla_{X}e_{1}=\omega(X)e_{2}\quad\nabla_{X}e_{2}=-\omega(X)e_{1}, where \omega(X)=\langle\nabla_{X}e_{1},e_{2}\rangle=-\langle\nabla_{X}e_{2},e_{1}\rangle . Then we have \left\{\begin{split}&\nabla_{e_{1}}e_{1}=\omega(e_{1})e_{2},\quad\nabla_{e_{1}}e_{2}=-\omega(e_{1})e_{1}\\ &\nabla_{e_{2}}e_{1}=\omega(e_{2})e_{2},\quad\nabla_{e_{2}}e_{2}=-\omega(e_{2})e_{1}.\end{split}\right. (2.3) Using the Codazzi equations, we have \left\{\begin{array}[]{ll}e_{2}(\kappa_{1})=(\kappa_{1}-\kappa_{2})\omega(e_{1})\\ e_{1}(\kappa_{2})=(\kappa_{1}-\kappa_{2})\omega(e_{2}),\end{array}\right. (2.4) where, as usually, e_{i}(f) denotes the derivative of a smooth function f along the vector e_{i} , i=1,2 . Notice that \kappa_{1}-\kappa_{2}\not=0 in \Omega . Thanks to ( 2.4 ), in the set \Omega , the Gauss equation gives the following expression of the Gauss curvature K , K=\kappa_{1}\kappa_{2}=-e_{1}\left(\frac{e_{1}(\kappa_{2})}{\kappa_{1}-\kappa_{2}}\right)+e_{2}\left(\frac{e_{2}(\kappa_{1})}{\kappa_{1}-\kappa_{2}}\right)-\frac{e_{1}(\kappa_{2})^{2}+e_{2}(\kappa_{1})^{2}}{(\kappa_{1}-\kappa_{2})^{2}}. (2.5) Denote by \Phi\colon\Sigma\to\mathbb{R}^{3} the immersion of a singular minimal surface. Let decompose the vector \vec{a} in its tangent part \vec{a}^{\top} and normal part \vec{a}^{\perp} with respect to \Sigma , \vec{a}^{\top}=\gamma e_{1}+\mu e_{2}, for some smooth functions \gamma and \mu . The functions \gamma and \mu are given by \gamma=\langle\vec{a},e_{1}\rangle,\quad\mu=\langle\vec{a},e_{2}\rangle. The normal part \vec{a}^{\perp} is \langle N,\vec{a}\rangle N , which by ( 1.1 ) is \vec{a}^{\perp}=\frac{H\langle\Phi,\vec{a}\rangle}{\alpha}N. Lemma 2.1 . Let \Omega be a subset of non-umbilic points of an \alpha -singular minimal surface. Then the functions \gamma and \mu satisfy the following equations \left\{\begin{split}0=&e_{1}(\gamma)-\mu\omega(e_{1})-\frac{H\langle\Phi,\vec{a}\rangle}{\alpha}\kappa_{1}\\ 0=&e_{2}(\gamma)-\mu\omega(e_{2})\\ 0=&e_{1}(\mu)+\gamma\omega(e_{1})\\ 0=&e_{2}(\mu)+\gamma\omega(e_{2})-\frac{H\langle\Phi,\vec{a}\rangle}{\alpha}\kappa_{2}\\ 0=&e_{1}(\frac{H\langle\Phi,\vec{a}\rangle}{\alpha})+\gamma\kappa_{1}\\ 0=&e_{2}(\frac{H\langle\Phi,\vec{a}\rangle}{\alpha})+\mu\kappa_{2}.\end{split}\right. (2.6) Proof. If X\in\mathfrak{X}(\Sigma) , then \overline{\nabla}_{X}\vec{a}=0 because \vec{a} is constant in \mathbb{R}^{3} . Then the Gauss and Weingarten formulas imply \begin{split}0&=\overline{\nabla}_{X}\vec{a}=\overline{\nabla}_{X}(\vec{a}^{\top}+\vec{a}^{\perp})=\overline{\nabla}_{X}\vec{a}^{\top}+\overline{\nabla}_{X}\vec{a}^{\perp}\\ &={\nabla}_{X}\vec{a}^{\top}+h(X,\vec{a}^{\top})+X(H\frac{\langle\Phi,\vec{a}\rangle}{\alpha})N-H\frac{\langle\Phi,\vec{a}\rangle}{\alpha}SX\end{split} for all X\in\mathfrak{X}(\Sigma) . Taking the tangent and the normal part in both sides of the above expression, we obtain \left\{\begin{array}[]{ll}{\nabla}_{X}\vec{a}^{\top}-H\frac{\langle\Phi,\vec{a}\rangle}{\alpha}SX=0\\ h(X,\vec{a}^{\top})+X(H\frac{\langle\Phi,\vec{a}\rangle}{\alpha})N=0.\end{array}\right. (2.7) Equations ( 2.6 ) are obtained by substituting in the two previous identities X by e_{1} and e_{2} and writing in coordinates with respect to \{e_{1},e_{2},N\} . Indeed, if X=e_{1} , and using ( 2.3 ), the first equation in ( 2.7 ) is \begin{array}[]{ll}0&={\nabla}_{e_{1}}(\gamma e_{1}+\mu e_{2})-H\frac{\langle\Phi,\vec{a}\rangle}{\alpha}\kappa_{1}e_{1}\\ &=e_{1}(\gamma)e_{1}+e_{1}(\mu)e_{2}+\gamma\omega(e_{1})e_{2}-\mu\omega(e_{1})e_{1}-H\frac{\langle\Phi,\vec{a}\rangle}{\alpha}\kappa_{1}e_{1}.\end{array} Similarly, the second equation in ( 2.7 ) is 0=h(e_{1},\gamma e_{1}+\mu e_{2})+e_{1}(H\frac{\langle\Phi,\vec{a}\rangle}{\alpha})N=\left(\gamma\kappa_{1}+e_{1}\left(\frac{H\langle\Phi,\vec{a}\rangle}{\alpha}\right)\right)N. This gives the first, third and fifth equations in ( 2.6 ). The other equations of ( 2.6 ) are obtained by putting X=e_{2} in ( 2.7 ). ∎ 3. Proof of Theorem 1.1 We know that planes parallel to \vec{a} , spheres centered at \Pi and cylindrical singular minimal surfaces have constant Gauss curvature. We prove the converse. The case

Read the complete text