1-efficient triangulations and the index of a cusped hyperbolic 3-manifold (Q888885)

If \(M\) is a three-manifold with nonempty toric boundary, an ideal triangulation of \(M\) is a triangulation \(\mathcal T\) such that \(M\) is homeomorphic to the underlying space minus a neighbourhood of each vertex. Such triangulations always exist; they are called 1-efficient if they do not have normal embedded spheres (or \(1\)-sided projective planes) or non-boundary parallel (in \(M\)) tori (or \(1\)-sided Klein bottles). Geometrisation implies that a manifold admitting a \(1\)-efficient triangulation is either hyperbolic or belongs to a small list of Seifert manifolds. It is proven here (following \textit{E. Kang} and \textit{J. H. Rubinstein} [Algebr. Geom. Topol. 5, 1505--1533 (2005; Zbl 1096.57018)]) that such triangulations can be endowed with an additional structure called an index structure (whose definition is too intricate to be detailed here, see Definition 2.12 in the paper). In turn an index structure allows to define the 3D-index of the triangulation, originally introduced by Dimofte, Gaiotto and Gukov [\textit{T. Dimofte} et al., Adv. Theor. Math. Phys. 17, No. 5, 975--1076 (2013; Zbl 1297.81149); Commun. Math. Phys. 325, No. 2, 367--419 (2014; Zbl 1292.57012)], which is a function associating to a choice of boundary curves a Laurent power series. It is not known whether the 3D-index is a topological invariant, i.e. whether it is independent of the ideal triangulation. In the present paper the authors associate to a hyperbolic manifold a collection of particular \(1\)-efficient triangulations which all have the same (well-defined) 3D-index. To prove this result they have to check that any two triangulations in this set are connected by a sequence of moves that do not affect the 3D-index. They use two such moves, the 2-3/3-2 moves and the 0-2/2-0 moves. The first is already known to preserve the 3D-index [\textit{S. Garoufalidis}, The 3D index of an ideal triangulation and angle structures, {\url arXiv:1208.1663}, to appear in Ramanujan math. journal) and they prove that the second one does as well. The triangulations are obtained via hyperbolic geometry, more precisely through the Epstein-Penner cell decomposition [\textit{D. B. A. Epstein} and \textit{R. C. Penner}, J. Differ. Geom. 27, No. 1, 67--80 (1988; Zbl 0611.53036)]. This decomposition is well-defined for a topological manifold, but complications arise due to the fact that its cells, though polyhedral, are not in general tetrahedra.