RICH bore radii should be parameterized

So far the bore radii are "guessed", to leave enough room for the beam lines. A better solution is to parameterize along an eta line (~3.7)