Hyperbolic Simplices of Maximal Inradius
Bruno DUCHESNE, Christopher-Lloyd SIMON · 2025
Read Hyperbolic Simplices of Maximal Inradius by Bruno DUCHESNE, Christopher-Lloyd SIMON (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-2512-17096.txt.
Introduction
Bruno Duchesne and Christopher-Lloyd Simon study hyperbolic simplices with maximal inradius, the largest ball that fits inside. A simplex here is set by n plus 1 points in compactified hyperbolic space, and its inradius is the Hausdorff gap between its full skeleton and its faces.
They show a simplex hits maximal inradius exactly when it is total ideal and regular, with a clean value whose tanh equals 1 over n. They also compute the maximal distance to the 1-skeleton, with tanh squared equal to n minus 1 over 2n. Those maxima stay uniformly bounded as dimension grows, tending to log of 1 plus root 2 for the 1-skeleton case.
Catalog record
| CREATOR | Bruno DUCHESNE, Christopher-Lloyd SIMON |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-17096
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Hyperbolic Simplices of Maximal Inradius Bruno DUCHESNE, Christopher-Lloyd SIMON Abstract For n ∈ ℕ n\in\mathbb{N} , consider a hyperbolic n n -dimensional simplex Δ \Delta , defined by 1 + n 1+n points in the compactified hyperbolic space 𝐇 n ⊔ ∂ 𝐇 n \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} . For each integer m ≤ n m\leq n , denote δ m n ( Δ ) ∈ [ 0 , + ∞ ] \delta^{n}_{m}(\Delta)\in[0,+\infty] the Hausdorff distance between its skeleta of dimensions n n and m m . In particular, δ n − 1 n ( Δ ) \delta^{n}_{n-1}(\Delta) is its inradius. The maximum of δ m n ( Δ ) \delta^{n}_{m}(\Delta) over Δ ∈ ( 𝐇 n ⊔ ∂ 𝐇 n ) 1 + n \Delta\in(\mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n})^{1+n} is denoted μ m n ∈ [ 0 , + ∞ ] \mu^{n}_{m}\in[0,+\infty] . We first show that Δ \Delta has maximal inradius δ n − 1 n ( Δ ) = μ m n \delta^{n}_{n-1}(\Delta)=\mu^{n}_{m} if and only if its is (total) ideal and regular; for which the inradius is given by tanh μ n − 1 n = 1 / n \tanh\mu^{n}_{n-1}=1/n . We deduce that Δ \Delta has maximal δ n − 1 n ( Δ ) = μ m n \delta^{n}_{n-1}(\Delta)=\mu^{n}_{m} if and only if it is (total) ideal and regular. We compute that the maximal distance to the 1 1 -skeleton μ 1 n \mu^{n}_{1} is given by ( tanh μ 1 n ) 2 = ( n − 1 ) / ( 2 n ) \left(\tanh\mu^{n}_{1}\right)^{2}=(n-1)/(2n) and deduce that those are uniformly bounded by lim n μ 1 n = log ( 1 + 2 ) \lim_{n}\mu^{n}_{1}=\log(1+\sqrt{2}) . 0 Introduction For n\in\mathbb{N}_{\geq 2} , let \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} be the compactified real algebraic hyperbolic space of dimension n . We will recall its definition and main properties in Section 1 . Here, a simplex \Delta is defined by a sequence of 1+n points in \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} called its vertices . It is total when its vertices are independent. It is ideal when all its vertices belong to \partial\mathbf{H}^{n} . For m\in\{0,\dots,n\} , a m -face of \Delta is the convex hull of 1+m vertices. Its m -skeleton \Delta^{(m)} is the union of its m -faces. For n>m>0 , we study the Hausdorff distance from \Delta^{(n)} to \Delta^{(m)} : \delta^{n}_{m}(\Delta)=\max\{d(p,\Delta^{(m)})\colon p\in\Delta^{(n)}\} . In particular \delta^{n}_{n-1}(\Delta) is the inradius of \Delta ( Section 2.2 ). Figure 1: Visualizing \delta^{3}_{2} and \delta^{3}_{1} in the incentred Euclidean model for an ideal simplex of \mathbf{H}^{3} . Section 2.2 introduces the incentred Euclidean model of a hyperbolic simplex, establishing a correspondence between ideal simplices and Euclidean simplices inscribed in the unit sphere whose incenter lies at the origin, which is equivariant under \Isom(\mathbf{H}^{n}) and \Isom(\mathbf{S}^{n-1}) . A total simplex of \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} is regular when its stabilizer in \Isom(\mathbf{H}^{n}) is isomorphic to the symmetric group \mathfrak{S}_{n+1} of its vertices. There exists a total ideal simplex of \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} which is regular, it is unique up to the action of \Isom(\mathbf{H}^{n}) . Theorem 0.1 (maximal inradius) . A total simplex \Delta of \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} has inradius \delta^{n}_{n-1}(\Delta)\leq\tanh^{-1}(1/n) , with equality if and only if it is ideal and regular. Proof outline. We work in the incentred Euclidean model, and use barycentric coordinates to show that Euclidean simplices with maximal inradius admit an orthocenter that coincides with the incenter. The regularity then follows from [ 5 , Theorem 4.3] . ∎ Remark 0.2 (maximal inradius) . We have no reference questioning or probing the characterization of hyperbolic simplices with maximal inradius, even though the inradius is closely studied or extensively used in several works [ 7 , 9 ] . Remark 0.3 (maximal volume) . The simplices of \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} with maximal volume are precisely those which are total ideal and regular. This was conjectured by Thurston and a complete proof was first given by [ 6 ] using analytic methods, then by [ 9 ] using Steiner symmetrization. We wonder if our Corollary 2.3 and the method in Theorem 2.6 could yield a simpler proof. For n,m\in\mathbb{N} , on the compact space (\mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n})^{n+1} of hyperbolic simplices in dimension n , the function \delta^{n}_{m} is continuous so it admits a maximum \mu^{n}_{m}=\max_{\Delta}\delta^{n}_{m}(\Delta)\in[0,\infty] , which must be achieved on a closed subset of total simplices which are ideal. Theorem 0.4 (maximal Hausdorff distance to m -skeleton) . For integers n>m>0 , a simplex \Delta in \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} satisfies \delta^{n}_{m}(\Delta)=\mu^{n}_{m} if and only if \Delta is total, ideal and regular. Moreover, \mu^{n}_{1} is given by \left(\tanh\mu^{n}_{1}\right)^{2}=\tfrac{n-1}{2n} so \mu^{n}_{1} grows as n\to\infty to \mu^{\infty}_{1}=\tanh^{-1}(1/\sqrt{2})=\log(1+\sqrt{2}) . Proof outline. For a total ideal simplex \Delta , we consider the successive projections of p\in\Delta^{(n-c)} on the closest faces of increasing codimension c : applying the hyperbolic Pythagorean theorem and the inequality in Theorem 0.1 shows that equality is achieved only for regular \Delta . ∎ Remark 0.5 (bounding \mu^{n}_{1} ) . The finiteness of \mu^{n}_{1} is mentioned in the proof of [ 1 , Theorem 3] with a reference to [ 2 ] , but our proof is complete and simple. In [ 1 , Theorem 3] , the finiteness of \mu^{n}_{1} is used to show certain results about the convergence of group actions on \mathbf{H}^{n} for the equivariant Gromov-Hausdorff topology. We were interested (and surprised) by fact that \mu^{n}_{1} is bounded by \mu^{\infty}_{1} , as this enabled us to extend these results in [ 3 ] and show Gromov-Hausdorff convergence for sequences of group actions on \mathbf{H}^{n_{i}} in a setting where n_{i}\in\mathbb{N}\cup\{\infty\} is any sequence of countable cardinals. We finish with the following 2.11 concerning the enumeration and localization of maximizers of \delta^{n}_{m}(\Delta) , and the Example 2.12 of disphenoids. Problem 0.6 (maximizers of \delta^{n}_{m} ) . Each (n+1) -set of m -faces \{F_{0}^{(m)},\dots,F_{n}^{(m)}\}\in\left(\genfrac{}{}{0.0pt}{}{\binom{\Delta}{1+m}}{1+n}\right) defines a decreasing intersection of convex sets in \Delta^{(n)} by C_{t}(F)=\{p\in\Delta^{(n)}\colon d(p,F_{i}^{(m)}\geq t\} , hence a point maximizing the distance to these faces, that is a local maximizer of the Hausdorff distance from \Delta^{(n)} to \Delta^{(m)} , which for some F will be a global maximer of \delta^{n}_{m} . What can be the number of distinct local global minimizers and where can they be located? Note that the isometry group of the simplex \operatorname{Stab}(\Delta^{(n)})\subset\Isom(\mathbf{H}^{n}) acts on that set of (1+n) -sets of m -faces. However the description and enumeration of distinct (local or global) maximizers is more subtle than the question of classifying (all or certain) of the orbits under this action, since it may happen that two (1+n) -sets of m -faces yield the same (local or global) maximizer without belonging to the same orbit. 1 Geometry of the algebraic hyperbolic space In this section, after recalling some geometry of the Lorentz-Minkowski space time, we define the compactified real algebraic hyperbolic space of dimension n\in\mathbb{N}_{>0} denoted \mathbf{H}^{n}\sqcup\partial\mathbf{H}^{n} by constructing three models, and use each of them to discuss some of its properties. The Minkowski hyperboloid model is suited to compute metric quantities (angles, distances and volumes) in \mathbf{H}^{n} using the algebra of the scalar product. The Cayley-Klein projective model is adapted for drawing and reasoning with incidence and orthogonality relations between points, lines and quadrics. We will use the Euclidean ball model to work around a distinguished point