Nonslice linear combinations of iterated torus knots (Q6164241)

A knot in the 3-sphere is algebraic if it is the link of an isolated singularity of a smooth curve. It is known that algebraic knots are certain iterated torus knots. Rudolph's conjecture, from the 1970's, states that algebraic knots are linearly independent in the smooth concordance group \(\mathcal{C}\). There is a lot of evidence for this conjecture. For example, \textit{R. A. Litherland} [Lect. Notes Math. 722, 71--84 (1979; Zbl 0412.57002)] proved that positive torus knots are linearly independent in \(\mathcal{C}\). Also, it follows from the work of \textit{K. Miyazaki} [Trans. Am. Math. Soc. 341, No. 1, 1--44 (1994; Zbl 0816.57007)], that the slice-ribbon conjecture implies Rudolph's conjecture. Furthermore, by work of \textit{T. Abe} and \textit{K. Tagami} [Math. Res. Lett. 23, No. 2, 303--323 (2016; Zbl 1357.57009)] and \textit{K. L. Baker} [J. Topol. 9, No. 1, 1--4 (2016; Zbl 1335.57010)], it follows that the slice-ribbon conjecture implies a stronger conjecture: The set of prime fibered strongly quasipositive knots is linearly independent in the smooth knot concordance group \(\mathcal{C}\). In this paper the authors consider families \(\mathcal{S}_p\) of iterated torus knots \(T(p,q_1;p,q_2;\dots;p,q_\ell)\), where \(p\) is a prime power and \((q_1,q_2,\dots,q_\ell)\) is a sequence of positive integers satisfying certain technical conditions. The main result of the paper shows that for every prime power \(p\), the set \(\mathcal{S}_p\) is linearly independent in the topological knot concordance group \(\mathcal{C}^{\mathrm{top}}\). It follows that for each prime power \(p\), the set \(\mathcal{S}_p^{\mathrm{alg}}\), consisting of the algebraic knots in \(\mathcal{S}_p\), satisfies Rudolph's conjecture. This set is in fact infinite. It also follows that for every prime power \(p\), the set \(\mathcal{S}_p\) satisfies the second conjecture, and that \(\mathcal{S}_p \backslash \mathcal{S}^{\mathrm{alg}}\) is an infinite family of nonalgebraic knots satisfying the second conjecture. The proof is technical, it depends on the metabelian Blanchfield pairings introduced by \textit{A. N. Miller} and \textit{M. Powell} [Algebr. Geom. Topol. 18, No. 6, 3425--3476 (2018; Zbl 1406.57005)].