Contactomorphic vertically convex domains

Jan Eyll, Jonas Fritsch, Kai Zehmisch · 2026

Read Contactomorphic vertically convex domains by Jan Eyll, Jonas Fritsch, Kai Zehmisch (2026) 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-2601-14842.txt.

Introduction

Jan Eyll, Jonas Fritsch, and Kai Zehmisch study shapes in contact geometry, a branch of math close to the study of motion. Their 2026 arXiv paper works in the standard Darboux space, a model space with a set contact form. The form they use is symmetric around a center ray.

The team looks at maps that keep the contact structure and fix the rim of a domain. They sort the domains that are straight in the vertical sense and round like a ball. Two such domains count as the same only under strict rules.

Catalog record

CREATORJan Eyll, Jonas Fritsch, Kai Zehmisch
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE87 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-14842

Browse the full library · About the Tower of Babel

Full text

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

Contactomorphic vertically convex domains Thanks: This work is part of a project in the SFB/TRR 191 Symplectic Structures in Geometry, Algebra and Dynamics , funded by the DFG Jan Eyll, Jonas Fritsch, Kai Zehmisch Abstract . We consider the standard Darboux space equipped with the radial symmetric contact form. We study co-orientation preserving contactomorphisms between relatively compact domains up to the boundary. We determine the contactomorphism classes among all strict vertically convex domains over a round ball in the Liouville hyperplane that are radially symmetric about the Reeb axis and whose boundary coincide along a neighbourhood of the common equator. The total invariant is the mean curvature of the bounding sphere at the umbilic points with the same sign. Replacing the Liouville hyperplane by codisc bundles of closed non-Besse Riemannian manifolds or finite symplectisations of closed non-Besse strict contact manifolds analogous results are formulated in terms of characteristic length and total characteristic action, resp. 2020 Mathematics Subject Classification 53D35; 53D05, 37C27, 37J55, 57R17. 1. Introduction Let V be a smooth manifold and consider a connected, compact hypersurface S in \mathbb{R}\times V . Assume that S bounds a (unique) relatively compact domain. The closure D_{S} is called the domain of S . We assume that S is strict vertically convex in the following sense: There exist smooth functions f_{\pm}\colon\thinspace V_{S}\rightarrow\mathbb{R} defined on an open subset V_{S} of V with f_{-}<f_{+} . Denote the graphs of f_{\pm} by S^{\pm} . Both functions f_{\pm} are assumed to extend over the boundary \partial V_{S} of V_{S} continuously with f_{-}=f_{+} along \partial V_{S} . We require S to be the (disjoint) union of the entrance set S^{-} , the exit set S^{+} and the graph of f_{\pm}|_{\partial V_{S}} , the so-called equatorial set \partial S^{\pm} of S . In case \partial S^{\pm} is a submanifold of S we will say equator for short and write E=\partial S^{\pm} . The triple (S,V_{S},f_{\pm}) is called a shape and V_{S} the shadow set of S . If V admits an exact symplectic form \mathrm{d}\lambda , a preferred choice of a Liouville primitive \lambda allows to form its contactisation (\mathbb{R}\times V,\mathrm{d}b+\lambda) together with the contact structure \xi_{\lambda}=\ker(\mathrm{d}b+\lambda) . A strict vertically convex hypersurface S in (\mathbb{R}\times V,\mathrm{d}b+\lambda) can be used to provide a boundary condition to compact contact manifolds with boundary. If a collar neighbourhood is required to be contactomorphic to a collar neighbourhood of S in (D_{S},\xi_{\lambda}) , the diffeomorphism type naturally compares to D_{S} . This was used in order to determine the diffeomorphism type under additional assumptions on the supported Reeb dynamics, see [ 2 , 3 , 6 , 8 , 9 , 10 , 12 ] . To come up with the contactomorphism type known tools are the classification of tight contact structures on the 3 -ball ( [ 5 ] ) and on a solid torus ( [ 14 ] ) in terms of the characteristic foliation on the boundary. In order to find examples for which the contactomorphism type of the domain of a strict vertically convex hypersurface can be determined, we will use the following notion: Two strict vertically convex hypersurfaces S and T in \mathbb{R}\times V are called to admit equivalent shapes (S,V_{S},f_{\pm}) and (T,V_{T},g_{\pm}) whenever V_{S}=V_{T} and the germs of continuous functions f_{+} and g_{+} as well as f_{-} and g_{-} , resp, coincide along \partial V_{S}=\partial V_{T} . In particular, S and T coincide in a neighbourhood of the common equatorial set. Consider the standard contactisation \big(\mathbb{R}\times\mathbb{R}^{2n},\mathrm{d}b+\tfrac{1}{2}(\mathbf{x}\mathrm{d}\mathbf{y}-\mathbf{y}\mathrm{d}\mathbf{x})\big) , whose contact structure is denoted by \xi_{\st} . A function f defined on an open subset of \mathbb{R}^{2n} is called radially symmetric provided that f is invariant under the action of the orthogonal group. We will identify f(\mathbf{x},\mathbf{y}) with f(r) for r=|(\mathbf{x},\mathbf{y})| . Theorem 1 . Let (S,V_{S},f_{\pm}) and (T,V_{T},g_{\pm}) be equivalent shapes in \mathbb{R}\times\mathbb{R}^{2n} with shadow set V_{S}=V_{T} equal to a round ball about 0 in \mathbb{R}^{2n} . Assume that f_{\pm} and g_{\pm} are all radially symmetric. Then the domains (D_{S},\xi_{\st}) and (D_{T},\xi_{\st}) are co-orientation preserving contactomorphic up to the boundary if and only if f^{\prime\prime}_{+}(0)=g^{\prime\prime}_{+}(0) and f^{\prime\prime}_{-}(0)=g^{\prime\prime}_{-}(0) . In other words, vertically convex domains with equivalent radially symmetric shapes are contactomorphic precisely if the mean curvatures at the umbilic points of the same sign coincide. The constructive part uses a Moser integrating vector field, so that in fact an ambient compactly supported contact isotopy relative to a neighbourhood of the equator is obtained. The induced isotopy of the boundary of the vertically convex domain is through radially symmetric shapes. In Theorem 4.7.1 and Section 4.8 we will consider symmetric shapes with shadow sets of more general Stein type. We will describe a perturbation method in order to obtain shapes with contactomorphic domains. To formulate an example we consider a closed, connected Riemannian manifold Q . The total space V=T^{*}Q of the cotangent bundle \pi\colon\thinspace T^{*}Q\rightarrow Q is provided with the canonical Liouville 1 -form \lambda=\lambda_{Q} , which at u\in T^{*}Q is given by u\circ T\pi . The contactisation \big(\mathbb{R}\times V,\mathrm{d}b+\lambda\big) is the 1 -jet bundle of Q with contact structure \ker(\mathrm{d}b+\lambda) denoted by \xi_{Q} . We call a smooth function f on an open subset of T^{*}Q fibred radially symmetric provided f(u) can be identified with f(r) for r=|u| w.r.t. the dual metric. Define the characteristic half length \mathscr{L}(f_{\pm}):=\pm\int_{0}^{r_{0}}\frac{f^{\prime}_{\pm}(\tau)}{\tau}\mathrm{d}\tau whenever f=f_{\pm} continuously extends to the closed co-disc bundle D_{r_{0}}T^{*}Q of radius r_{0}>0 and defines the entrance or exit set of a strict vertically convex hypersurface S in \mathbb{R}\times V via the graph. Belonging to the exit/entrance set determines the sign \pm . The quantity \mathscr{L}(f_{\pm}) is derived from the characteristic foliation on the graph of f_{\pm} in \big(\mathbb{R}\times V,\mathrm{d}b+\lambda\big) . Namely |\mathscr{L}(f_{\pm})| is the length of a geodesic arc that is obtained from the projection of a half characteristic to Q by deleting consecutive equal length subarcs of opposite orientation, see Sections 3.1 and 5.5 . Define the characteristic length to be \Sigma(S):=\mathscr{L}(f_{+})+\mathscr{L}(f_{-})\,. Theorem 2 . Let (S,V_{S},f_{\pm}) and (T,V_{T},g_{\pm}) be equivalent shapes in \mathbb{R}\times T^{*}Q with shadow set V_{S}=V_{T} equal to the interior of D_{r_{0}}T^{*}Q . Assume that f_{\pm} and g_{\pm} are fibred radially symmetric and that \Sigma(S)=\Sigma(T) . Then the domains (D_{S},\xi_{Q}) and (D_{T},\xi_{Q}) are co-orientation preserving contactomorphic up to the boundary. Moreover, the existing contactomorphism can be assumed (a) to restrict to \id\times\,\eta_{\theta_{0}} in a neighbourhood of the equator, where \eta_{\theta_{0}} is the normalised cogeodesic flow at time \theta_{0} , and (b) to map \{f_{\pm}(0)\}\times Q to \{g_{\pm}(0)\}\times Q via \big(f_{\pm}(0),q\big)\mapsto\big(g_{\pm}(0),q\big) , resp. Theorem 2 is based on a Moser argument analogous to the if part of Theorem 1 after potentially applying a normalised cogeodesic twist considered in Section 5.6 . Observe, that the reparametrisation \eta_{\theta/r_{0}} is the Reeb flow at time \theta/r_{0} on the co-sphere bundle S_{r_{0}}T^{*}Q with respect to the contact form \alpha^{r_{0}} that is obtained from restricting \lambda_{Q} to T(S_{r_{0}}T^{*}Q) . The proof of Theorem 2 shows that if the equality \mathscr{L}(f_{\pm})=\mathscr{L}(g_{\pm}) holds for the characteristic half length, resp., the time \theta_{0} in Theorem 2 (a) can be chosen to be zero, so that the resulting contactomorphism fixes the equator pointwise. To formulate a converse statement to Theorem 2 notice that by Section 3.5 any contactomorphism (D_{S},\xi_{Q})\rightarrow(D_{T},\xi_{Q}) can be assumed to preserve

Read the complete text