Dear motorway,
I'd a quick scan through your structure, so I'll just give a few opinions.
Is that a monopole or a dipole you're simulating? Cos' the hf$$ structure looks like a quartered-up (you're making use of the image planes, right?) coax-fed monopole with curvy ground plain.
So, with respect to a pretty decent monopole simulation in HF$$, I think there are a few changes to be made:
(1) You don't have to use image-planes. It's a simple structure, you don't have to complicate your simulation results for that.
Besides, I think you gotta do some post-processing (e.g. multiply the simulated impedance results for the final one) if you use the image-thingy.
(2) You don't have to have such a long coax-feed. Roughly the same height as your coax dielectric's diameter will do. This was suggested to me by HF$$ engineers.
(3) I think you need larger air-box/radiation boundaries. Your radiation boundaries should be at least 0.25 wavelengths (or more) to the tip(s) of your radiating element.
Hope these helps,
Dave