Log in to the online community
(as one of my many hats here is tech authority for our HV test faclilty.)
Actually on test rigs that are a bit near the knuckle, a ball of blue tak over sharp screws etc to round off to larger radius of curvature on a a semi-insulating surface can be quite handy, Not for final production of course..
Actually for a given geometry, negative corona sets in at a noticeably lower voltage gradient - electrons are easier to pull off a metal than out of the air. (or if you like electrons are runnier than holes...)
With that in mind, we normally do DC test structures at negative voltages, if it holds up, it will normally also pass the positive case with no effort.