Is it normal to get S11=-4dB when simulating patch antenna in Sonnet Lite?

I am curious aboout S11 value at center frequency... is it normal to get such a low value of s11? I think that my antenna must have s11 value of around -15dB (simulation and measurement results from one paper on patch antennas) but instead i have only -4dB. I am using adaptive sweep. Top box is open. It seems all is setup correctly. Anybody had a such problem?

Something might be wrong with your setup, or the feedpoint. In any case, you can get much better matching if antenna and simulatuion setup are correct. Can you zip and upload the *.son project file?

I tried to make single microstrip patch projects from scratch. It looks nice with 5Ghz, 10Ghz and 12GHz single patch (lambda/4 matched). But when i go at higher freq for example 26GHz, S11 result of adaptive sweep looks ugly: no smoothing, and some "spikes" on the graph.

This might be box resonances in your Sonnet simulation box. The Sonnet method needs metal walls at the sides, and this isn't ideal for antenna simulation because it might cause resonances (standing waves) inside the simulation box. In the xgeom editor menu, use Analysis > Estimate Box Resonances to check for box resonances. Use Circuit > Box and make sure that your top cover of the simulation is set to "Free Space".