The BiGGEST MWS problem competition

Hi all EM fans!
Is anybody interested in large problems in MWS?
I have one to start with :)
-Parabolic antenna with hyperbolic subreflector, dimensions of whole objects are approx. 32.6 lambda x 32.6 lambda x 18.6 lambda
-Electric and magnetic wall symmetry used (so only 1/4 of the structure is meshed).
-Operating frequency is 35GHz and parabola diameter is 28cm.
-6.6 millions of cells (12 cells per lambda)
- approx. 1GB of RAM taken by mws calculation..
Please post your tasks here too 8O
Rgz
eirp

Well, my simulation turned to problem
The results aren't as expected..
- Calc. directivity is 23dBi, expected 30+ dBi
- Someting strange is at the mainlobe (twisted like)
- Parabola illumination is wrong
Who can comment this?
What's possible:
- error in first geometrical optic design
- mistake in parabola input curve
- usage of PEC and PMC wall in MWS
Thanks in advance to all antenna gurus!
Rgz
eirp

Hi eirp,
Can you post or send me the model so I can have a look at it.
BR,
jjantenna

Dear all!
Problem solved (big thanks to goxy and others)
There was mistake, the feedhorn + subreflector was in incorrect distance from the parabola :)
Calculated dir is 36.9 dBi!
Now all is fine!
Cheers
eirp

Hi eirp,
Anyway we could get the model files (.sat) or a script for this interesting antenna problem? I think many could benefit from another MWS "example".
Thanks for your consideration.
regards,
MoonShine

It's better to get xxx.mod and xxx.par then the sat file
they are small if you want the proj.
RGZ
PL.

Here is the project.
You need good machine with at least 1GB of RAM.
Cheers
eirp
PS: Feeding horn needs small modification to improve the dish behavior in side lobes