Small volume link orbifolds (Q2511478)

The unique orientable hyperbolic \(3\)-orbifold of smallest volume was identified by \textit{F. W. Gehring} and \textit{G. J. Martin} [Ann. Math. (2) 170, No. 1, 123--161 (2009; Zbl 1171.30014)], \textit{T. H. Marshall} and \textit{G. J. Martin} [ibid. 176, No. 1, 261--301 (2012; Zbl 1252.30030)]. In the present paper, the authors approach the same issue restricting to the family of link orbifolds, i.e. \(3\)-orbifolds with a link as singular locus. Based on computer experiments and on results in [\textit{F. W. Gehring} and \textit{G. J. Martin}, loc. cit.], they conjecture that the unique hyperbolic link orbifold of smallest volume is \(\mathcal{O}_L\), a two-fold quotient of the knot orbifold \(\mathcal{O}_K\) whose singular locus is the \(5_2\) knot with torsion order 3. In the first result of the paper the authors restrict to the subclass of orbifolds with a \(\mathbb{Z}_6\) homology sphere as base space, and provide a lower bound to the volume of a link orbifold in such a class. Moreover, they show that in such a class the smallest volume hyperbolic knot (resp. link) orbifold is \(\mathcal{O}_K\) (resp. \(\mathcal{O}_L\)). As first step in the proof, it is shown that the minimal volume link orbifold has torsion orders only \(2\) and/or \(3\) and as base space a closed \(3\)-manifold. In view of this fact the authors restrict their analysis to orbifolds satisfying such properties. For orbifolds with only 2-torsion, using the homological hypothesis of the theorem, the authors construct a 2-fold manifold cover with hyperbolic structure, and similarly they provide a 3-fold manifold cover for orbifolds with no 2-torsion. Hence the volume of the orbifold is bounded from below by one half or one third of the volume of the Weeks manifold (the unique closed, orientable, hyperbolic \(3\)-manifold of minimal volume). Since the Weeks manifold is a three-fold cyclic orbifold cover of \(\mathcal{O}_K\), the result follows. In the presence of both torsion orders 2 and 3, an ad hoc combination of the constructions developed in the two previous cases yields the conclusion. The second part of the paper is devoted to relaxing the homological hypothesis of the main result obtaining weaker volume estimates. Inspired by the work of \textit{D. Gabai} et al. [J. Am. Math. Soc. 22, No. 4, 1157--1215 (2009; Zbl 1204.57013)]; [\textit{P. Milley}, J. Topol. 2, No. 1, 181--192 (2009; Zbl 1165.57016)], the authors drill out the singular locus of an orbifold \(\mathcal{O}\) obtaining a cusped manifold \(M_\mathcal{O}\), and hence recover the initial orbifold \(\mathcal{O}\) by doing Dehn filling. The computation of the volume of the initial orbifold relies on the estimation of the variation of volume at each step. More precisely, they deduce a version for orbifolds of the work by \textit{I. Agol} et al. [J. Am. Math. Soc. 20, No. 4, 1053--1077 (2007; Zbl 1155.58016)] for estimating the change of volume produced by the drilling operation. The volume of \(M_\mathcal{O}\) is controlled using a result by D. Gabai et al. [loc. cit.]. Hence, the variation of volume originated by the Dehn filling is estimated using a mild generalization of a theorem by \textit{D. Futer} et al. [J. Differ. Geom. 78, No. 3, 429--464 (2008; Zbl 1144.57014)]. This procedure is applied in all but finitely many cases which are detected using computer programs Tube and Snap. Finally, with similar techniques they identify the minimal volume hyperbolic link orbifold without any 2- or 3-torsion.