A triangulation of CP ^3 as symmetric cube of S ^2

From MaRDI portal
Publication:452007



Abstract: The symmetric group S3 acts on S2imesS2imesS2 by coordinate permutation, and the quotient space (S2imesS2imesS2)/S3 is homeomorphic to the complex projective space CCP3. In this paper, we construct an 124-vertex simplicial subdivision (S2imesS2imesS2)124 of the 64-vertex standard cellulation S42imesS42imesS42 of S2imesS2imesS2, such that the S3-action on this cellulation naturally extends to an action on (S2imesS2imesS2)124. Further, the S3-action on (S2imesS2imesS2)124 is "good", so that the quotient simplicial complex (S2imesS2imesS2)124/S3 is a 30-vertex triangulation CCP303 of CCP3. In other words, we construct a simplicial realization (S2imesS2imesS2)124oCCP303 of the branched covering S2imesS2imesS2oCCP3. Finally, we apply the BISTELLAR program of Lutz on CCP303, resulting in an 18-vertex 2-neighbourly triangulation CCP183 of CCP3. The automorphism group of CCP183 is trivial. It may be recalled that, by a result of Arnoux and Marin, any triangulation of CCP3 requires at least 17 vertices. So, CCP183 is close to vertex-minimal, if not actually vertex-minimal. Moreover, no explicit triangulation of CCP3 was known so far.


In a previous paper [ibid. 46, 542--560 (2011; Zbl 1231.51016)] the authors constructed a 10-vertex triangulation of \(\mathbb{C} P^2\) as a branched quotient of the 16-vertex product \(S^2\times S^2\), subdivided in a certain way. This branched simplicial covering preserves a large part of the symmetry. In the paper under review, they start with the 64-vertex product \(S^2\times S^2\times S^2\), subdivided in a certain way, and construct a 30-vertex simplicial triangulation of \(\mathbb{C}^3\) as a branched quotient. Here the subdivision requires 60 extra vertices since there is no simultaneous and coherent triangulation of all the various cubes that come in here. The mapping from \((S^2)^3\) to \(\mathbb{C} P^3\) is a simplicial branched covering, and again a large symmetry group is preserved. By bi-stellar moves, the number of vertices for \(\mathbb{C} P^3\) could be reduced to 18 but not (or not yet) to 17. There is no 16-vertex triangulation according to \textit{P. Arnoux} and \textit{A. Marin} [Mem. Fac. Sci., Kyushu Univ., Ser. A 45, No. 2, 1670--244 (1991; Zbl 0753.52002)].





Describes a project that uses

Uses Software






This page was built for publication: A triangulation of \(\mathbb CP ^{3}\) as symmetric cube of \(S ^{2}\)

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q452007)