In mathematics, hyperbolic Dehn surgery is an operation by which one can obtain further hyperbolic 3-manifolds from a given cusped hyperbolic 3-manifold. Hyperbolic Dehn surgery exists only in dimension three and is one which distinguishes hyperbolic geometry in three dimensions from other dimensions.
Such an operation is often also called hyperbolic Dehn filling, as Dehn surgery proper refers to a "drill and fill" operation on a link which consists of drilling out a neighborhood of the link and then filling back in with solid tori. Hyperbolic Dehn surgery actually only involves "filling".
We will generally assume that a hyperbolic 3-manifold is complete. Suppose M is a cusped hyperbolic 3-manifold with n cusps. M can be thought of, topologically, as the interior of a compact manifold with toral boundary. Suppose we have chosen a meridian and longitude for each boundary torus, i.e. simple closed curves that are generators for the fundamental group of the torus. Let
M(u1,u2,...,un)
ui=pi/qi
pi
qi
ui
infty
M(infty,...,infty)
We equip the space H of finite volume hyperbolic 3-manifolds with the geometric topology.
The Thurston's hyperbolic Dehn surgery theorem states
M(u1,u2,...,un)
Ei
M(u1,u2,...,un)
2 | |
p | |
i |
→ infty
pi/qi
ui
\omega\omega
The figure-eight knot and the (-2, 3, 7) pretzel knot are the only two knots whose complements are known to have more than 6 exceptional surgeries; they have 10 and 7, respectively. Cameron Gordon conjectured that 10 is the largest possible number of exceptional surgeries of any hyperbolic knot complement. This was proved by Marc Lackenby and Rob Meyerhoff, who show that the number of exceptional slopes is 10 for any compact orientable 3-manifold with boundary a torus and interior finite-volume hyperbolic. Their proof relies on the proof of the geometrization conjecture originated by Grigori Perelman and on computer assistance. It is currently unknown whether the figure-eight knot is the only one that achieves the bound of 10. One conjecture is that the bound (except for the two knots mentioned) is 6. Agol has shown that there are only finitely many cases in which the number of exceptional slopes is 9 or 10.