A complete system of inequalities for the diameter, in-, and circumradius in the 3 3 -dimensional Euclidean space

René Brandenberg, Bernardo González Merino, Mia Runge · 2025

Read A complete system of inequalities for the diameter, in-, and circumradius in the 3 3 -dimensional Euclidean space by René Brandenberg, Bernardo González Merino, Mia Runge (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-2509-05028.txt.

Introduction

Three numbers describe a solid's size from different angles: its diameter, its inradius, and its circumradius. René Brandenberg, lead author with two co-authors, present a complete system of inequalities linking those quantities in three-dimensional Euclidean space in a 2025 geometry paper. A complete system means the inequalities capture all relations among the three measures.

To build it, they prove quasiconcavity of the inradius over simplices that share a common facet. That result holds independently of the norm or gauge used to measure distance. The paper finishes a classical problem about how the three size measures constrain one another.

Catalog record

CREATORRené Brandenberg, Bernardo González Merino, Mia Runge
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE55 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-05028

Browse the full library · About the Tower of Babel

Full text

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

A complete system of inequalities for the diameter, in-, and circumradius in the 3 3 -dimensional Euclidean space Thanks: 2020 Mathematics Subject Classification. 52A40 ; 52A15 . This research has been funded by the PID2022-136320NB-I00 project / AEI/10.13039/501100011033/ FEDER, UE René Brandenberg, Bernardo González Merino, Mia Runge Abstract . We present a complete system of inequalities for the inradius, circumradius, and diameter in the 3 3 -dimensional Euclidean space. To do so, we prove quasiconcavity of the inradius evaluated over n n -simplices with a common facet independently of the norm/gauge under consideration. 1. Introduction Let \mathds{B}_{2}^{n} denote the n -dimensional (Euclidean) unit ball and \mathcal{K}^{n} be the set of convex bodies (i.e. non-empty, convex, and compact sets) in \mathds{R}^{n} . For any K\in\mathcal{K}^{n} , let R(K) be the circumradius of K , (i.e. the smallest \rho\geq 0 such that a translation of a ball of radius \rho covers K ) and r(K) be the inradius of K (i.e. the largest \rho\geq 0 such that a translation of a ball of radius \rho is contained in K ). Finally, let D(K) be the diameter of K (i.e. the maximal length of a segment within K ). The aim of this paper is to describe the range of values that the inradius, circumradius and diameter of K in the 3 -dimensional Euclidean space may achieve. To do so, we compute a complete system of inequalities for those functionals, i.e. a list of inequalities such that if and only if a given 3-tuple of parameters (r,D,R) fulfills all those inequalities, there exists a convex body whose inradius, diameter, and circumradius coincide with those parameters. Theorem 1.1 . Let K\in\mathcal{K}^{3} . Then, (1) 2R(K)\geq D(K),\quad\sqrt{3}D(K)\geq\sqrt{8}R(K),\quad D(K)\geq r(K)+R(K),\quad r(K)\geq 0, and whenever D(K)\leq\sqrt{3}R(K) holds true then (2) r(K)\geq\frac{D(K)^{2}\sqrt{3R(K)^{2}-D(K)^{2}}}{4R(K)\sqrt{3R(K)^{2}-D(K)^{2}}+\sqrt{3}(4R(K)^{2}-D(K)^{2})}\ . Moreover, ( 1 ) and ( 2 ) state a complete system of inequalities for the inradius, diameter, and circumradius in Euclidean 3-space. The new contribution within this theorem is the last inequality. We prove the validity of ( 2 ) and also that it is sharp if and only if K is a 3-simplex having at least five diametrical edges (which includes the equilateral 3-simplex) or K is an equilateral triangle. A lower bound for the inradius has been derived by Dekster forty years ago [ 7 , Lem. 1.3 (5)] , including the open task of finding a best possible inequality [ 7 , Rem. (1)] . 1 1 1 ( 2 ) has been explicitly determined (and claimed valid and optimal by experimental observations) within the Master thesis “Complete Systems of Inequalities Describing the Feasible Configurations of Triples of Geometrical Functionals of M. Horsch, at Technical University of Munich (2019).” This is the very first time that such a complete system of inequalities for a triple of functionals has been derived for the whole family \mathcal{K}^{3} of 3 -dimensional convex bodies. So far, only in [ 13 ] , restricting to centrally symmetric convex bodies, such a system could be derived. In order to visualize Theorem 1.1 , we consider the mapping (3) f:\mathcal{K}^{3}\rightarrow[0,1]^{2},\quad\quad f(K):=\left(\frac{r(K)}{R(K)},\frac{D(K)}{2R(K)}\right). Realize that a complete description of the so-called (r,D,R) -Blaschke-Santaló diagram f(\mathcal{K}^{3}) is equivalent to providing a complete system of inequalities for the inradius, circumradius and diameter. See Figure 1 for a sketch of the (r,D,R) -diagram ( 3 ). Historically, it was Blaschke who, in 1916, first proposed the question of what values the volume, surface area, and integral mean curvature of three-dimensional convex bodies can have [ 1 ] . Later on Santaló [ 16 ] studied complete systems of inequalities for planar sets for triples of geometric functionals (including the planar analog of Theorem 1.1 ). Several other authors continued Santaló’s work [ 3 , 8 , 11 , 12 ] and even different functionals [ 9 , 10 ] or four functionals at the same time [ 4 , 18 ] have been considered. In order to show the fifth inequality in Theorem 1.1 , we prove a quasiconcavity property for the inradius with respect to simplices sharing a common facet. We do so, not restricting to the Euclidean case. For K,C\in\mathcal{K}^{n} , let r(K,C) denote the inradius of K with respect to C , i.e. the largest \rho\geq 0 such that a translation of K contains \rho C . The set of extreme points of K , i. e. those points p in K that are not contained in the convex hull of K\setminus\left\{p\right\} , is denoted by \ext(K) . Theorem 1.2 . Let C\in\mathcal{K}^{n} be full-dimensional and p^{2},\ldots,p^{n+1}\in\mathds{R}^{n} affinely independent, and P be a convex set contained in the open half-space bounded by the affine hull of p^{2},\ldots,p^{n+1} . Define S_{p} , p\in P , to be the simplex with vertices \left\{p,p^{2},\ldots,p^{n+1}\right\} . Then there exists p^{*}\in\ext(P) such that S_{p^{*}} has minimal inradius r(S_{p^{*}},C) over all simplices S_{p},p\in P . Let us observe that Theorem 1.2 has been proven for the 2 -dimensional Euclidean case [ 16 ] , but never before in its full generality. This may open the door for obtaining new inequalities in higher dimensions, for different gauges, and other combinations of functionals. The paper is organized as follows. In Section 2 , we collect the definitions and technical results needed throughout the paper. In Section 3 , we prove Theorem 1.2 and in Section 4 , we show the new inequality ( 2 ) in order to prove Theorem 1.1 . \frac{r}{R} \frac{D}{2R} \phantom{\sqrt{\frac{8}{3}}}1\phantom{\sqrt{\frac{8}{3}}} \phantom{\sqrt{\frac{8}{3}}}\frac{1}{3}\phantom{\sqrt{\frac{8}{3}}} 1 \sqrt{\frac{8}{3}}-1 \frac{\sqrt{3}}{2} \frac{2}{\sqrt{6}} Figure 1. The new diagram f(\mathcal{K}^{3}) . 2. Technical results and definitions For any X\subset\mathds{R}^{n} , the linear, affine, and convex hull are denoted by \lin(X) , \aff(X) , and \conv(X) , respectively. The convex hull of two points x and y is called a segment and is abbreviated by [x,y] . The open segment is denoted by (x,y) . The convex hull of n+1 affinely independent points is called a simplex . The boundary of X is described by \bd(X) and the interior by \inte(X) . Analogously, the relative boundary \relbd(X) and relative interior \relint(X) of X are the boundary and interior of X evaluated within \aff(X) . For any X,Y\subset\mathds{R}^{n} and \rho\in\mathds{R} let X+Y:=\{x+y:x\in X,y\in Y\} be the Minkowski sum of X and Y and \rho X:=\{\rho x:x\in X\} the \rho - dilatation of X . We abbreviate \left\{x\right\}+Y=:x+Y and (-1)X=:-X . The support function of K\in\mathcal{K}^{n} is defined as h_{K}(\cdot):\mathds{R}^{n}\to\mathds{R} , h_{K}(a):=\max_{x\in K}a^{T}x and the polar as K^{\circ}:=\{x\in\mathds{R}^{n}:x^{T}y\leq h_{C}(y)\text{ for all }y\in K\} . The circumradius of K\in\mathcal{K}^{n} is defined as R(K):=\min\{\rho\geq 0:\exists t\in\mathds{R}^{n}\text{ such that }K\subset t+\rho\mathds{B}^{n}_{2}\} and the diameter as the longest segment in K : D(K):=\max_{x,y\in K}\left\|x-y\right\|_{2}. The inradius of K\in\mathcal{K}^{n} with respect to C\in\mathcal{K}^{n} is defined as r(K,C):=\max\{\rho\geq 0:\exists t\in\mathds{R}^{n}\text{ such that }t+\rho C\subset K\} and we abbreviate r(K):=r(K,\mathds{B}^{n}_{2}) . A set t+r(K,C)C , t\in\mathds{R}^{n} , which is contained in K is called an inball of K . One may recognize that R(K)=R(\ext(K)) and D(K)=D(\ext(K)) (c.f. [ 2 ] ), which in particular means that the diameter of a polytope, and more specifically a simplex, is attained between two of its vertices. One of the first inequalities found, relating these functionals, is Jung’s inequality, bounding the diameter from below by the

Read the complete text