Normal forms of functions with degenerate singularities on surfaces equipped with semi-free circle actions

Bohdan Feshchenko · 2024

Read Normal forms of functions with degenerate singularities on surfaces equipped with semi-free circle actions by Bohdan Feshchenko (2024) 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-2412-18944.txt.

Introduction

Bohdan Feshchenko studies a class of smooth functions on surfaces in a 2024 math paper. The functions take circle values and live on shapes like a cylinder, a torus, a disk, or a sphere. They widen a classic family that has no saddle points, and they respect a spinning symmetry of the surface.

Feshchenko proves a normal form for the whole class. Each such function can be built from the simplest Morse function on its surface, reshaped by a smooth warp of the surface and a smooth warp of the values. The result gives one tidy mold for a family that looks varied at first glance.

Catalog record

CREATORBohdan Feshchenko
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE61 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-18944

Browse the full library · About the Tower of Babel

Full text

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

Normal forms of functions with degenerate singularities on surfaces equipped with semi-free circle actions Bohdan Feshchenko Abstract . This article is devoted to the study of a certain class of smooth circle-valued functions on a cylinder S 1 × [ 0 , 1 ] S^{1}\times[0,1] , a torus T 2 T^{2} , a disk D 2 D^{2} and a sphere S 2 S^{2} which is a generalization of Morse-Bott functions without saddles. We established a "normal form" for functions from this class, namely, we proved that any such function represents in the form f = ϰ ∘ f 0 ∘ h − 1 f=\varkappa\circ f_{0}\circ h^{-1} , where f 0 f_{0} is the simplest Morse function on the given surface for some diffeomorphism h h of M M and a smooth function ϰ \varkappa satisfying some natural conditions. 1. Introduction The study of smooth functions on manifolds and their properties lies at the heart of a modern differential topology and a theory of dynamical systems. It connects local geometric properties of functions to the global topology of the underlying surfaces. This connection is well understood for generic classes of smooth functions like Morse functions, functions with isolated singularities, Morse-Bott functions. Let M be a smooth manifold, and P be either a real line \mathbb{R} or a circle S^{1} . Recall that two smooth P -valued function f,g:M\to P are called topologically (smoothly) equivalent, if there exist homeomorphisms (diffeomorphisms) h:M\to M and \ell:P\to P such that (1) f=\ell\circ g\circ h^{-1}. The question of when two functions are smoothly or topologically equivalent is challenging and remains open in general. Topological equivalence of functions with isolated singularities on smooth compact surfaces were studied by V. Sharko [ 17 ] , E. V. Kulinich [ 5 ] , A. Prishlyak [ 16 ] , A. Kadubovskiy [ 3 ] , for simple Morse functions by J. Martínez-Alfaro, I. S. Meza-Sarmiento, and R. Oliveira [ 13 ] , for circle-valued simple Morse-Bott functions were studied by E. B. Batista, J. C. F. Costa and I. S. Meza-Sarmiento [ 1 ] . V. Sharko obtained the description of connected components of the space of Morse functions. The functions (Definition 1.4 ) we will study in this paper are closely related to questions of the study of homotopy properties stabilizers of smooth functions. We will recall the necessary definitions. The product \mathcal{D}(P)\times\mathcal{D}(M) of the groups of diffeomorphisms of P and diffeomorphisms of M naturally acts from the left on the space of smooth P -valued functions C^{\infty}(M,P) by the rule: \gamma_{LR}:\mathcal{D}(P)\times\mathcal{D}(M)\times C^{\infty}(M,P)\to C^{\infty}(M,P),\qquad\gamma_{LR}(\ell,h,f)=\ell\circ f\circ h^{-1}. The group \mathcal{D}(M)=\mathrm{id}_{M}\times\mathcal{D}(M) is a subgroup of \mathcal{D}(P)\times\mathcal{D}(M) and \gamma_{LR} induces a left action \gamma_{R}:\mathcal{D}(M)\times C^{\infty}(M,P)\to C^{\infty}(M,P),\qquad\gamma_{R}(h,f)=f\circ h^{-1}. Actions \gamma_{LR} and \gamma_{R} are left actions, but it will be convenient to use the terms “left-right” and “right” to refer the sides at which we apply the corresponding diffeomorphisms. Note that two smooth functions f,g:M\to P are smooth equivalent, if they belongs to the same orbit of the action \gamma_{LR}. 1.1. Stabilizers of functions from C^{\infty}(M,P) Denote by \mathcal{D}^{+}(P) the group of orientation preserving diffeomorphisms of P. Let f be a smooth function in C^{\infty}(M,P) . The image \mathrm{Im}(f)=f(M) is a smooth submanifold of P . Denote by \mathcal{D}^{+}(\mathrm{Im}(f)) a subgroup of \mathcal{D}^{+}(P) . One can define stabilizers of f with respect to the restriction of the action \gamma_{LR} on \mathcal{D}^{+}(\mathrm{Im}(f))\times\mathcal{D}(M) and the action \gamma_{R} by the formulas: \displaystyle\mathcal{S}_{LR}(f) \displaystyle=\{(\ell,h)\in\mathcal{D}^{+}(\mathrm{Im}(f))\times\mathcal{D}(M)\,|\,\ell\circ f\circ h^{-1}=f\}, \displaystyle\mathcal{S}_{R}(f) \displaystyle=\{h\in\mathcal{D}(M)\,|\,f\circ h^{-1}=f\}. There is a canonical inclusion j:\mathcal{S}_{R}(f)\hookrightarrow\mathcal{S}_{LR}(f) given by j(h)=(\mathrm{id}_{\mathrm{Im}(f)},h) . Denote by \mathcal{S}_{R}^{\mathrm{id}}(f) a connected component of \mathcal{S}_{R}(f) containing \mathrm{id}_{M}. Homotopy properties of these spaces and their connected components are well studied for some “wide” class smooth functions with isolated singularities on surfaces by S. Maksymenko, see [ 9 , 11 ] and also a review paper [ 12 ] . In [ 2 ] we studied such questions on stabilizers (see Theorem 1.6 below) for a broader class of smooth functions on surfaces with non-isolated singularities \mathcal{F}(M,P) , whose connected components of the set of critical points of f are either isolated critical points, or critical circles: Definition 1.2 (cf. Definition 1.1 [ 2 ] ) . Let M be compact surface possible with boundary \partial M , and \mathcal{F}(M,P) be a class of smooth functions on M which satisfy: (1) a function f|_{\partial M} is locally constant, (2) a set of critical points \Sigma_{f} of f is a disjoint union of smooth submanifolds of M and \Sigma_{f}\subset\mathrm{Int}(M), (3) for each connected component C of \Sigma_{f} and each critical point p\in C there exist a local chart (U,\phi:U\to\mathbb{R}^{2}) near p with \phi(p)=0 and a chart (V,\psi:V\to\mathbb{R}) near f(p)\in P such that f(U)\subset V and a local representation f_{p}=\psi\circ f\circ\phi^{-1}:\phi(U)\to\psi(V) of f is (3.a) either a polynomial homogeneous polynomial f_{p} without multiple factors, (3.b) or is given by f_{p}(x,y)=\pm y^{n_{C}} for some n_{C}\in\mathbb{N}_{\geq 2} depending of C. It is known that for f\in\mathcal{F}(M,P) the stabilizer \mathcal{S}_{R}(f) is strong deformation retract of \mathcal{S}_{LR}(f) , so the inclusion j:\mathcal{S}_{R}(f)\hookrightarrow\mathcal{S}_{LR}(f) as above is a homotopy equivalence, see [ 7 , Theorem 1.3] . Note that recently we established the homotopy type of the connected component \mathcal{S}^{\mathrm{id}}_{R}(f) of the identity map in \mathcal{S}_{R}(f) for functions from \mathcal{F}(M,P) , see Theorem 1.6 below. 1.3. Functions from \mathcal{F}^{\circ}(M,P) and their normal forms Consider the following class subclass \mathcal{F}^{\circ}(M,P) of \mathcal{F}(M,P) : Definition 1.4 . Assume that M is an oriented surface. Denote by \mathcal{F}^{\circ}(M,P) a class of functions which satisfy (1), (2), (3.b) of Definition 1.2 , and (3.a^{\prime}) instead of (3.a): ( 3.a^{\prime} ) either a polynomial f_{p}:\mathbb{R}^{2}\to\mathbb{R} , given by f_{p}(x,y)=\pm(x^{2}+y^{2}) . Generalities of function from \mathcal{F}^{\circ}(M,P) will be discussed in Section 2 , but here we only mention that the class \mathcal{F}^{\circ}(M,P) is “small” in the sense that functions from \mathcal{F}^{\circ}(M,P) appear on several “simple” surfaces: Proposition 1.5 . Let f be a function from \mathcal{F}^{\circ}(M,P) . (1) Then M is diffeomorphic to one of the following four surfaces: a cylinder S^{1}\times[0,1] , a disk D^{2}, a sphere S^{2} or a torus T^{2} . (2) If M is diffeomorphic to • a cylinder or a torus, then f has no isolated critical points, • a disk, then f has one Morse extremum (minimum or maximum), • a sphere then f has two Morse extrema (minimum and/or maximum). Moreover, all four cases are possible (depends on the number of critical circles). Notice that surfaces from Proposition 1.5 admit smooth semi-free actions of a circle S^{1} . Our motivation to study functions from \mathcal{F}^{\circ}(M,P) is the following result on the homotopy type of \mathcal{S}_{R}^{\mathrm{id}}(f) for functions from \mathcal{F}(M,P) mentioned above. Theorem 1.6 (Theorem 1.2 [ 2 ] , cf. Theorem 1.3 [ 9 ] ) . Let f be a function from \mathcal{F}(M,P) . Then \mathcal{S}_{R}^{\mathrm{id}}(f) is either contractible or homotopy equivalent to S^{1} . In particular \mathcal{S}_{R}^{\mathrm{id}}(f)\simeq\begin{cases}S^{1},&\text{if $M$ is oriented and $f\in\mathcal{F}^{\circ}(M,P)$},\\ \mathrm{pt},&\text{otherwise}.\end{cases} Our aim is to give an “analytic” description of all functions from this class, see Theorem 1.8 . There are some trivial examples of functions from \mathcal{F}^{\circ}(M,P) that are easy to write by hand: Definition 1.7 . Let f_{0}:M_{0}\to P be a smooth function from \mathcal{F}^{\circ} ( 1_{0} ) M_{0}=S^{1}\times[0,1]=\{(z,s)\,|\,z\in\mathbb{C},|z|=1,\,0\leq s\leq 1\} is a unit cylinder, and f_{0}:S^{1}\times[0,1]\to\mathbb{R} is

Read the complete text