Special curves and Hamiltonian systems from polynomial eigenfunctions of an elliptic operator
Juan García Escudero · 2025
Read Special curves and Hamiltonian systems from polynomial eigenfunctions of an elliptic operator by Juan García Escudero (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-2508-17065.txt.
Introduction
Juan García Escudero studies a family of polynomials tied to a classic three-cusped curve, the deltoid, and the surfaces built from its tangent lines. Earlier work laid out the polynomials and their algebraic surfaces. This paper follows two threads: special plane curves linked to the family, and dynamical systems whose energy functions come from the same polynomials. For those Hamiltonian systems, the author analyzes the bifurcations, where the motion changes character as conditions vary.
Catalog record
| CREATOR | Juan García Escudero |
|---|---|
| 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/2508-17065
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Special curves and Hamiltonian systems from polynomial eigenfunctions of an elliptic operator Juan García Escudero Abstract . A family of polynomials linked to the set of the deltoid tangents and its associated algebraic hypersurfaces has been presented in recent years. In this paper we study some related maximising and free plane curves. We also analyse the bifurcations on polynomial Hamiltonian dynamical systems defined from such a family. keywords singularities, special plane curves, polynomial dynamical systems 1991 Mathematics Subject Classification 14B05, 14H50, 35J15, 37J38 1. Introduction We study some plane curves and Hamiltonian systems related to a family of polynomials introduced in the last decade for the construction of algebraic hypersurfaces with many simple singularities ( [ 8 , 9 , 10 , 13 ] and references within). The one-parameter family of degree d polynomials P_{d}(x,y,\mu) is discussed in Section 2, where we also consider the connection of P_{d}(x,y,\mu) with eigenfunctions of a symmetric diffusion operator (see [ 13 ] for a generalisation to n variables) which corresponds to a Laplace-Beltrami operator for a metric with zero scalar curvature. Special curves called free and maximising plane curves are significant in the field of algebraic geometry. Maximising curves can be used for instance in the construction of algebraic surfaces having the maximal Picard number. A link between maximising curves and free curves has been found in [ 5 ] . In Section 3 we study some free and maximising curves related to P_{d}(x,y,\mu) . In particular we discuss maximising curves of degrees eight and ten. A degree nine curve which is free but not maximising is presented. We also show that a certain maximising decic has a Miyaoka-Kobayashi number close to the maximum possible. The critical points of P_{d}(x,y,\mu) and some of their properties are presented in Section 4. We use them in Section 5 in order to analyse the bifurcation or catastrophe sets of the associated polynomial Hamiltonian dynamical systems. We find six bifurcation points if d is divisible by three and two bifurcation points otherwise. 2. A family of degree d polynomials as eigenfunctions of an elliptic operator We write P_{d}(x,y,\mu) instead of the polynomials \hat{J}_{d,\tau=\mu}(x,y) introduced in [ 10 ] . Up to a scaling factor, the polynomials P_{d}(x,y,\mu) are connected to the union of the lines with d orientations linked with the affine Weyl group of the root lattice \mathbb{A}_{2} and the set of the deltoid tangents [ 12 ] . The lines have equations L_{d,\mu,s}(x,y):=y+({\rm cos}2\varphi-x){\rm tan}\varphi+{\rm sin}2\varphi=0,\varphi=\frac{6s-1}{6d}-\frac{\mu}{\pi d},s=-\lfloor\frac{d-2}{2}\rfloor,-\lfloor\frac{d-2}{2}\rfloor+1,...,\lfloor\frac{d+1}{2}\rfloor with (x,y)\in{\mathbb{R}}^{2} and \mu\in\mathbb{R} . The basic degree- d polynomials are P_{d}(x,y,\mu):=\lambda_{d,\mu}\prod_{s}L_{d,\mu,s}(x,y) where \lambda_{d,\mu}=(-1)^{m}2d if \mu=(6m-3d-1)\frac{\pi}{6} (in this case the line L_{d,\mu,s}(x,y)=0 parallel to the y -axis is interpreted as the line x+1=0 ) and \lambda_{d,\mu}=2{\rm cos}(\mu+\frac{d\pi}{2}+\frac{2\pi}{3}) if \mu\neq(6m-3d-1)\frac{\pi}{6} , m\in\mathbb{Z} . Symmetric diffusion operators defined on an open subset \Omega\subset{\mathbb{R}}^{n} can be written [ 3 ] {\mathcal{L}}:=\sum_{ij}g^{ij}(x)\partial_{x_{i}x_{j}}^{2}+\sum_{i}b^{i}(x)\partial_{x_{i}} (2.1) such that the symmetric matrix (g^{ij}(x)) is non negative at every point x=(x_{1},x_{2},...x_{n})\in\Omega . {\mathcal{L}} is elliptic in the interior of \Omega and an operator of this kind in {\mathbb{R}}^{2} , denoted by {\mathcal{L}}_{2} in [ 10 ] , has g^{11}(x_{1},x_{2})=-\frac{1}{4}(3x_{1}^{2}-x_{2}^{2}-6x_{1}-9),g^{12}(x_{1},x_{2})=g^{21}(x_{1},x_{2})=-\frac{1}{2}(2x_{1}x_{2}+3x_{2}), g^{22}(x_{1},x_{2})=-\frac{1}{4}(3x_{2}^{2}-x_{1}^{2}+6x_{1}-9),b^{1}(x_{1},x_{2})=-x_{1},b^{2}(x_{1},x_{2})=-x_{2} (2.2) The operator {\mathcal{L}}_{2} is elliptic in the interior of the region -Q_{\delta}(x_{1},x_{2})\geq 0 , where Q_{\delta}(x_{1},x_{2})=(x_{1}^{2}+x_{2}^{2})^{2}-8(x_{1}^{3}-3x_{1}x_{2}^{2})+18(x_{1}^{2}+x_{2}^{2})-27=0 is the equation of a Steiner hypocycloid or deltoid \delta . For each \mu , J_{d}(x_{1},x_{2},\mu):=P_{d}(x_{1},x_{2},\mu)-2{\rm cos}3\mu is an eigenfunction of {\mathcal{L}}_{2} [ 10 ] : {\mathcal{L}}_{2}J_{d}(x_{1},x_{2},\mu)=-d^{2}J_{d}(x_{1},x_{2},\mu) (2.3) {\mathcal{L}}_{2} has the form of the Laplace-Beltrami operator \frac{1}{\sqrt{|g|}}\sum_{i,j}\partial_{x_{i}}(\sqrt{|g|}g^{ij}\partial_{x_{j}}) where g^{ij} is the (co)metric or inverse metric given in Eq. (2.2). The metric is g_{ij}=(g^{ij})^{-1} and |g|=\det g_{ij}=-\frac{3}{16}Q_{\delta} . By computing the Christoffel symbols \Gamma^{m}_{jk}:=\frac{1}{2}\sum_{i}g^{im}(\partial_{x_{k}}g_{ij}+\partial_{x_{j}}g_{ki}-\partial_{x_{i}}g_{jk}) and the Riemann curvature tensor R^{m}_{ijk}:=\partial_{x_{j}}\Gamma^{m}_{ki}-\partial_{x_{k}}\Gamma^{m}_{ij}+\sum_{s}\Gamma^{s}_{ki}\Gamma^{m}_{js}-\sum_{s}\Gamma^{s}_{ij}\Gamma^{m}_{ks} we obtain R^{m}_{ijk}=0 , hence also both the Ricci tensor R_{ik}:=\sum_{j}R^{j}_{ijk} and the scalar curvature R:=\sum_{i,k}g^{ik}R_{ik} are zero. The operator {\mathcal{L}}_{2} can be interpreted as a projection of the Euclidean Laplacian \Delta_{{\mathbb{R}}^{2}} . We use the operator {\mathcal{G}}_{{\mathcal{L}}}(f,g):=\frac{1}{2}({\mathcal{L}}(fg)-f{\mathcal{L}}g-g{\mathcal{L}}f) (2.4) and we see that g^{jk}={\mathcal{G}}_{{\mathcal{L}}}(x_{j},x_{k}),b^{j}={\mathcal{L}}x_{j} . In terms of z=x_{1}+\sqrt{-1}x_{2} and \bar{z}=x_{1}-\sqrt{-1}x_{2} the operator {\mathcal{L}}_{2} has a simpler form, denoted by {\mathcal{L}}_{2C} , with [ 10 ] {\mathcal{G}}_{2C}(z,z)=-z^{2}+3\bar{z},{\mathcal{G}}_{2C}(z,\bar{z})={\mathcal{G}}_{2C}(\bar{z},z)=9-z\bar{z},{\mathcal{G}}_{2C}(\bar{z},\bar{z})=-\bar{z}^{2}+3z,{\mathcal{L}}_{2C}z=-z,{\mathcal{L}}_{2C}\bar{z}=-\bar{z} (2.5) Under a change of variables y=(y_{k}(x))_{k=1,2,...} we have in general {\mathcal{L}}\phi(y)=\sum_{ij}{\mathcal{G}}_{{\mathcal{L}}}(y_{i},y_{j})\partial_{y_{i}y_{j}}^{2}\phi(y)+\sum_{i}({\mathcal{L}}y_{i})\partial_{y_{i}}\phi(y) (2.6) The so-called generalised cosine (see [ 9 ] and references therein) h(u,v):=e^{-2\pi u\sqrt{-1}}+e^{-2\pi v\sqrt{-1}}+e^{2\pi(u+v)\sqrt{-1}} (2.7) can be used to make a change of variables which shows that {\mathcal{L}}_{2} is a projection of \Delta_{{\mathbb{R}}^{2}} . We identify the sets {\mathbb{R}}^{2} and {\mathbb{C}} and we make the substitution 2\pi u=-\omega\cdot z,2\pi v=-{\bar{\omega}}\cdot z in Eq.(2.7), where z=x_{1}+\sqrt{-1}x_{2}\in{\mathbb{C}} , z_{1}\cdot z_{2}={\rm Re}(z_{1}\bar{z}_{2}) and 1,\omega,\bar{\omega} are the cubic roots of unity. The variable change Z:{\mathbb{R}}^{2}\rightarrow{\rm Int}\delta , where {\rm Int}\delta denotes the interior of the region whose border is the deltoid \delta , is Z(z)=e^{(1\cdot z)\sqrt{-1}}+e^{(\omega\cdot z)\sqrt{-1}}+e^{({\bar{\omega}}\cdot z)\sqrt{-1}} (2.8) It can been shown that \Delta_{{\mathbb{R}}^{2}}Z=-Z,\Delta_{{\mathbb{R}}^{2}}\bar{Z}=-\bar{Z} and {\mathcal{G}}_{\Delta_{{\mathbb{R}}^{2}}}={\mathcal{G}}_{2C} where {\mathcal{G}}_{2C} is given by Eq.(2.5) for z=Z,\bar{z}=\bar{Z} , hence {\mathcal{L}}_{2C}f(Z,\bar{Z})=\Delta_{{\mathbb{R}}^{2}}f(Z,\bar{Z}) (2.9) The Jacobian |\frac{\partial(x_{1}^{{{}^{\prime}}},x_{2}^{{{}^{\prime}}})}{\partial(x_{1},x_{2})}|=4\sqrt{3}({\rm cos}(3x_{1})-{\rm cos}(\sqrt{3}x_{2})){\rm sin}(\sqrt{3}x_{2}) , where x_{1}^{{{}^{\prime}}}={\rm Re}Z(z),x_{2}^{{{}^{\prime}}}={\rm Im}Z(z) , is zero in three families of lines, with equations 3x_{1}=\pm\sqrt{3}x_{2}+2k\pi,\sqrt{3}x_{2}=l\pi,k,l\in{\mathbb{Z}} , forming a periodic triangular tiling in {\mathbb{R}}^{2} . Z is a one-to-one map from the border of a triangle, which is a fundamental region of the affine Weyl group \mathbb{A}_{2} , to \delta . Also Z is a one-to-one map from the interior of the triangle onto {\rm Int}\delta and maps {\mathbb{R}}^{2} onto {\rm Int}\delta . 3. Some free and maximising curves In this Section we first recall some basic facts on free and maximising curves in \mathbb{P}^{2} and then we present some examples based on the polynomials P_{d}(x,y,\mu) . We also comment on specific curves with high Miyaoka-Kobayashi number. Let {\rm Der}(S)=\{\partial:=a\cdot\partial_{x}+b\cdot\partial_{y}+c\cdot\partial_{z},a,b,c\in S\} be the free S -module of {\mathbb{C}} -linear derivations of the polynomial ring S . The graded S -module of derivations for a reduced curve C:F=0 preserving the ideal \langle F\rangle is denoted by {\rm D}(F)=\{\partial\in{\rm Der}(S):\partial F\in\langle F\rangle\} . If \delta_{E}=x\partial_{x}+y\partial_{y}+z\partial_{z} is the Euler derivation and {\rm D}_{0}(F)=\{\partial\in{\rm Der}(S):\partial F=0\} is the graded S -module of Jacobian syzygies of F , then we have {\rm D}(F)={\rm D}_{0}(F)\oplus S\cdot\delta_{E} . Definition 3.1 . [ 6 ] We say that a reduced curve C:F=0 is free if {\rm D}(F) , or equivalently {\rm D}_{0}(F) , is a free graded S -module. The exponents (d_{1},d_{2}) of a free curve C are the degrees of a basis for the graded S -module {\rm D}_{0}(F) with rank 2. If C is a curve of even degree d with only simple singularities then the “number” of simple singularities (by counting each simple singularity X_{k},X=A,D,E by its index k ) of C is less or equal \frac{3}{4}d^{2}-\frac{3}{2}d+1 [ 16 ] . Now we use these results to construct algebraic curves with many simple singularities of types A,D and E [ 8 , 10 ] . The singularities have local equations [ 2 ] A_{k}:x^{k+1}+y^{2}=0;D_{k}:x(x^{k-2}+y^{2})=0;E_{6}:x^{3}+y^{4}=0;E_{7}:x(x^{2}+y^{3})=0 In [ 5 ] it is shown that reduced plane curves C of odd degree d=2m+1 with only ADE singularities have a total Tjurina number at most \tau(C)=3m^{2}+1 . Curves with maximum total Tjurina number of C , denoted by \tau(C) , were studied in the 80s ( [ 16 ] , [ 20 ] ). Later studies have motivated the following Definition 3.2 . [ 6 ] A curve C:F=0 of degree d having only ADE-singularities is maximising if either