Knot concordance and torsion (Q700492): Difference between revisions

It is an interesting and difficult question whether there exists a knot with finite order other than two in the classical knot concordance group. Note that in the algebraic concordance group and in higher odd dimensions, there are infinitely many such knots [\textit{J. Levine}, Comment. Math. Helv. 44, 229-244 (1969; Zbl 0176.22101), Invent. Math. 8, 98-110; addendum, ibid. 8, 355 (1969; Zbl 0179.52401)]. In this paper, the authors obtain a new condition for a knot to have an infinite order. A useful corollary to their main theorem is as follows: Let \(K\) be a knot and \(M_{K}\) be the double branched covering space of the three-sphere branched along \(K\). Suppose that the Alexander polynomial of \(K\) is \(nt^2-(2n+1)t+n\) such that the \(p\)-primary summand of \(H_{1}(M_{K})\) is cyclic, where \(n\) is a positive integer and \(p\) is a prime congruent to 3 modulo 4 such that \(4n+1=p^k q\) with \(k\) odd and \(q\) coprime to \(p\). Then \(K\) is of infinite order in the classical knot concordance group. Note that such a knot has order 4 in the algebraic concordance group. For the proof they use the Casson-Gordon invariant and the linking form of the double branched covering [\textit{A. J. Casson} and \textit{C. McA. Gordon}, Proc. Symp. Pure Math. Vol. 32, Part 2, 39-53 (1978; Zbl 0394.57008); Cobordism of classical knots. With an appendix by P. M. Gilmer. Progr. Math., 62, A la recherche de la topologie perdue, 181-199, Birkhäuser Boston, Boston, MA, 1986, see Zbl 0597.57001].