Hfss bug?Different angle, different result.

Hi, maybe I find a bug in Hfss.
A patch on the outside surface of a cylinder, at first, the resonance frequency is 2.45G, then rotate the patch about the z-axis 12 degree, and get a different resonance frequency :2.20G. And I find when the angle is 45,90, 135 and so on, the resonance frequency is 2.45G.

Desktop.zip

I have used the coax instead of the port in this model，but result is the same. I really can't understand it. If it is a bug, is anybody can give me a hand?