4
$\begingroup$

Let $T_1$ be the standart trefoil knot, embedded in $\mathbb R^3$. Then, one can easily give a simple Wirtinger presentation of $\pi_1(\mathbb R^3 \setminus T_1)$ by $\langle a,b,c | a = bcb^{-1}, b=cac^{-1}, c=aba^{-1} \rangle$. Killing one generator and substituting the corresponding, superflous relation into the other relators, we get the reduced presentation $\langle a,b| a = baba^{-1}b^{-1}, b= abab^{-1}a^{-1} \rangle$. Knowing that $\pi_1(\mathbb R^3 \setminus T_1)$ is isomorphic to the braid group $B_3$, I want to simplify it even further, so that $\pi_1(\mathbb R^3 \setminus T_1) = \langle x,y|x^3 = y^2 \rangle$. Now two obvious candidates are $x = ba$ and $y= bab$. Unsing the above relations, one can immediately compute that $a = yx^{-1}$ and $x^3=y^2$. However, I was so far unable to express $b$ in terms of $x$ and $y$. Can anybody help me out here, it might be something really obvious I am missing....

Further, let $T_1,....T_n$ be a collection of disjoint trefoils with $n \geq 2$ in $\mathbb R^3$ that also have disjoint images under a regular projection of the corresponding link. Then, one can construct the connected sum $T := T_1 \# .....\# T_n$, which is now again a knot in $\mathbb R^3$. Then, one can compute a group presentation of $\pi_1(\mathbb R^3 \setminus T)$ in terms of $n + 1$ generators and $n$ relators, given by $\langle x_i,y| x_i^3 = (x_i x_1^{-1}yx_1^{-1} x_i)^2\rangle $ for $i=1,..n$.

Question: Does there exist a presentation of $\pi_1(\mathbb R^3 \setminus T)$ with less than $n+1$ generators?

EDIT: I found a somewhat simpler presentation, given by $\langle a,b_i: ab_ia = b_iab_i \rangle$ for $i=1,...,n$. Is it clear that this presentation cannot be reduced further ?

It might very well be that a solution to the first question is very helpful for answering the second one. Any help is appreciated

$\endgroup$
2
  • 1
    $\begingroup$ $$b=(ba)^{-1}(ba)b=(ba)^{-1}(bab)=x^{-1}y.$$ $\endgroup$ Commented Jan 8, 2016 at 14:57
  • $\begingroup$ My mistake was that i was staring at the relations for too long. Thanks! $\endgroup$
    – H1ghfiv3
    Commented Jan 8, 2016 at 15:05

1 Answer 1

1
$\begingroup$

Richard Weidmann proved in his paper "On the rank of amalgamated products and product knot groups" that if $\mathcal K$ is a knot that is the connected sum of $n$ (non-trivial) knots, then $rank(\pi_1(\mathbb S^3 \setminus \mathcal K)) \geq n+1$. In my case, this implies that $rank(\pi_1(\mathbb R^3 \setminus T)) = n+1$.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .