Bi-contact structures with symmetry: local normal forms
Connor Jackman · 2025
Read Bi-contact structures with symmetry: local normal forms by Connor Jackman (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-2508-00802.txt.
Introduction
Connor Jackman studies pairs of contact structures that share a symmetry in a 2025 mathematics paper. A contact structure is a field of planes twisting through a three-dimensional space. Take two that cross transversally, and in general no flow preserves both at once. Jackman asks what such pairs look like when a family of symmetries does exist. He determines local normal forms, standard local shapes, for them. One case involves Anosov flows, which arise globally as the crossing of two oppositely oriented contact structures with maximal local symmetries.
Catalog record
| CREATOR | Connor Jackman |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 69 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-00802
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Bi-contact structures with symmetry: local normal forms Connor Jackman Abstract . A pair of transverse contact distributions on a 3-manifold will in general admit no 1-parameter families of symmetries: a flow preserving both contact distributions. Here, we will determine local normal forms for such pairs admitting symmetries. In particular, we observe that orientable Anosov flows may be globally given by the intersection of a pair of oppositely oriented contact distributions admitting, around any point, maximal local symmetries. 1. Introduction Pairs of transverse contact distributions on a 3-manifold have been an active topic of interest primarily due to their relations to projective Anosov flows (see §2.2 of [ 4 ] , and §4 of [ 12 ] ), and in particular Anosov flows (eg [ 8 ] , [ 10 ] ). Here we examine such pairs admitting infinitesimal symmetries, where: Definition 1 . A contact distribution, \xi , on a 3-manifold M is a non-integrable rank two distribution on M . We write (M,\xi,\tilde{\xi}) for a pair of transverse contact distributions \xi,\tilde{\xi} on the 3-manifold M . A symmetry of the pair is a diffeomorphism \varphi:M\to M preserving both contact distributions: \varphi_{*}\xi=\xi,\penalty\ \penalty\ \penalty\ \varphi_{*}\tilde{\xi}=\tilde{\xi}. A infinitesimal symmetry of the pair is a vector field whose flow is by symmetries of the pair. Definition 2 . We call the local symmetry algebra, \mathfrak{g} , of (M,\xi,\tilde{\xi}) around a point m\in M , the Lie algebra of infinitesimal symmetries of the pair defined in some neighborhood of m\in M . Note that the local symmetry algebra around a point is a real vector space, having a Lie algebra structure by Lie bracket of vector fields. As is well known, a single contact distribution on a 3-manifold admits many infinitesimal symmetries: all are locally equivalent to the standard Darboux normal form: \ker\{dy-pdx\},\penalty\ (x,y,p)\in\mathbb{R}^{3} , and any function H(x,y,p) generates an infinitesimal symmetry via its contact Hamiltonian vector field, eg: (1) H_{p}\partial_{x}+(pH_{p}-H)\partial_{y}-(H_{x}+pH_{y})\partial_{p} where here, and in what follows, we write subscripts for partial derivatives ( H_{p}=\partial_{p}H , etc.). On the other hand, as we will see, a typical pair of transverse contact distributions will admit no infinitesimal symmetries. As for those pairs of transverse contact distributions admitting infinitesimal symmetries, we find here the following local normal forms: Theorem 1.1 . Suppose a pair of transverse contact distributions \xi,\tilde{\xi} on a 3-manifold M admits an infinitesimal symmetry around m\in M transverse to \xi\cap\tilde{\xi} . Then around m , the pair is given in some coordinates by one of the following local normal forms: (I 1 ) For c\in\mathbb{R}\backslash\{0,1\} a constant: \xi=\{dy=pdx\},\penalty\ \penalty\ \penalty\ \tilde{\xi}=\{dy=cpdx\}. The symmetry algebra of such a pair is given by the vector fields of the form: \mathfrak{g}=\langle u(x)\partial_{x}+v(y)\partial_{y}+p(v^{\prime}(y)-u^{\prime}(x))\partial_{p}\rangle where u(x),v(y) are any smooth functions of x and y respectively, and u^{\prime},v^{\prime} their derivatives. (I 2 ) For c^{2}\leq 4 a constant, \xi=\{dy=pdx\},\penalty\ \penalty\ \penalty\ \tilde{\xi}=\{(p+c)dy+dx=0\}. The symmetry algebra of such a pair in these coordinates are those vector fields of the form: \mathfrak{g}=\langle u\partial_{x}+v\partial_{y}-u_{y}(1+cp+p^{2})\partial_{p}\rangle where u(x,y) is any solution to u_{xx}+u_{yy}=cu_{xy} and v(x,y) its conjugate solution determined, up to addition of a constant, through: dv=-u_{y}dx+(u_{x}-cu_{y})dy. (II 1 ) For f(y) a function: \xi=\{dy=pdx\},\penalty\ \penalty\ \tilde{\xi}=\{(p+f(y))dy\pm dx=0\},\penalty\ \penalty\ \penalty\ \mathfrak{g}=\langle\partial_{x}\rangle. (II 2 ) For f(y) a function: \xi=\{dy=pdx\},\penalty\ \penalty\ \penalty\ \tilde{\xi}=\{dy=f(y)pdx\},\penalty\ \penalty\ \penalty\ \mathfrak{g}=\langle u(x)\partial_{x}-pu^{\prime}(x)\partial_{p}\rangle. In the symmetry algebra, u(x) is any smooth function of x and u^{\prime} its derivative. (III 1 ) For functions a(y),b(y),f(p) : \xi=\{dy=\frac{p+a(y)}{p+b(y)}dx\},\penalty\ \penalty\ \tilde{\xi}=\{dy=\frac{f(p)+a(y)}{f(p)+b(y)}dx\},\penalty\ \penalty\ \penalty\ \mathfrak{g}=\langle\partial_{x}\rangle. (III 2 ) For a function f(p) : \xi=\{dy=pdx\},\penalty\ \penalty\ \penalty\ \tilde{\xi}=\{dy=f(p)dx\},\penalty\ \penalty\ \penalty\ \mathfrak{g}=\langle\partial_{x},\partial_{y},x\partial_{x}+y\partial_{y}\rangle. (IV) For a function f(y,p) : \xi=\{dy=pdx\},\penalty\ \penalty\ \tilde{\xi}=\{dy=f(y,p)dx\},\penalty\ \penalty\ \penalty\ \mathfrak{g}=\langle\partial_{x}\rangle. Remark 1 . These normal forms are determined up to re-labelings of the contact structures, and the open set containing m\in M is identified in each case with an appropriate open set U\subset\mathbb{R}^{3} (namely one where \xi,\tilde{\xi} are transverse contact distributions in these coordinates). For example in case I 1 , a neighborhood of m\in M is identified with an open subset U\subset\{p\neq 0\}\subset\mathbb{R}^{3} , or for I 2 , with U an open subset of the origin x=y=p=0 . In the functionally dependent families II - IV, the functions are not arbitrary but subject to certain restrictions in order to define transverse contact structures on some open set. Explicitly, for II 1 we may take a neighborhood U of the origin in \mathbb{R}^{3} , for II 2 : we can take U some neighborhood of x=y=0,p=1 for f(y)\neq 0,1 for y near 0. For types III we may take U as a neighborhood of the origin, where we have a(y)\neq b(y) for y near zero and f(p)\neq p having no fixed points for p near zero and f^{\prime}(p)\neq 0 for p near zero. For type IV, we may take some neighborhood of x=y=0,p=1 where we have f(y,p)\neq p and f_{p}(y,p)\neq 0 around y=0,p=1 . Remark 2 (On the transversal condition to the axis.) . For (M,\xi,\tilde{\xi}) admitting a non-trivial (non-vanishing) infinitesimal symmetry Y , the locus where Y\in\xi\cap\tilde{\xi} is tangent to this ‘axis’, \xi\cap\tilde{\xi} , is of codimension at least one (and in general of codimension two, locally given by an isolated integral curve of this axis) being the intersection of the hypersurfaces \{Y\in\xi\}\cap\{Y\in\tilde{\xi}\} (as Y\neq 0 these are locally hypersurfaces as follows from ( 1 )). It is possible, in certain cases, that the hypersurfaces \{Y\in\xi\} and \{Y\in\tilde{\xi}\} coincide, for example: \xi=\{dy=pdx\},\penalty\ \tilde{\xi}=\{dy=\pm pdx+dp\} and Y=\partial_{x} . Here we will be focused on normal forms around ‘regular’ points: where such a symmetry field is transverse to this axis \xi\cap\tilde{\xi} . It would be natural in further work to examine further local normal forms around such ‘singular’ points, for example: where a pair admits a symmetry tangent to \xi\cap\tilde{\xi} at said point, or around a point at which a symmetry field vanishes. Remark 3 . We would also like to note that our choice of coordinates in which we present the local normal forms above is somewhat arbitrary: only determined up to arbitrary diffeomorphisms. The forms presented in the above theorem 1.1 are merely those which appeared to us to have the ‘simplest’ defining formulas. The items in theorem 1.1 can be distinguished by considering certain local invariants associated to a pair (M,\xi,\tilde{\xi}) in order to be certain the items I–IV can be realized by truly distinct pairs (not equivalent under any local diffeomorphism). See table 1 below, for a summary of these conditions (our labelling into the types I – IV being motivated by the different types of invariants involved in this table), as well as remark 9 below for some of their simpler distinguishing properties. The paper is structured as follows: in § 2 we will define some basic invariants of a pair (M,\xi,\tilde{\xi}) of transverse contact structures, and in § 4 carry out the relevant computations to establish the normal forms stated in the main theorem 1.1 , introducing the additional invariants which may be used to distinguish them along the way. In § 3 we will give a pair of examples in which the normal forms of items I 1 (with c=-1 ) and I 2 (with c=2 ) are realized: relating to Anosov flows and integrable contact flows respectively. 2. Invariants We define some basic