Designing dielectric profiles to shape electric fields and/or inhibit either air breakdown or surface tracking (two entirely different things) does take some care and background knowledge. If your approach is one of trial and error then you can punch embarrassing holes in lots of pieces of plastic before you find a design which works, so it's worth choosing a material which is cheap to buy and quick and easy to machine.
Seams are a particular problem. Were you hoping to join the two halves of your annulus with a seam which would be as electrically strong as the bulk material ? If you're careful then it is possible to make high-dielectric-strength epoxy joints to glass-reinforced plastic. But you need to make sure the surfaces are clean and relatively flat as you can't afford to have significant dielectric discontinuities (these can act as field concentrators).
The AWE pulsed power group became wonderfully capable at building extremely high-voltage equipment quickly and inexpensively over a period stretching from the 1950s to at least the 1980s. I was lucky enough to learn a lot from them when I was a student and then an early researcher. Their preferred insulator was acrylic (Perspex) joined using a relatively quick-setting two-part compound called Simplex Rapid (a liquid mixed with a powder - we always used the ivory coloured powder although I'm not sure whether the colour made any difference). This stuff was developed for use mostly by dental technicians
http://www.kemdent.co.uk/simplex-rapid-liquid-1l. But it was electrically pretty strong too, and quite resistant to surface tracking.
Are you sure you will actually need a guard ? Quite a lot can be achieved to reduce tracking and, in particular, corona using spray-on conformal coating
http://cpc.farnell.com/ambersil/6130...SABEgKVuPD_BwE.
Cheers,
GJ