help,problems about HFSS9

I use HFSS9 to simulate EBG structure.But everytime it tells me that the result can't be

converged,and I don't know there's the problem.here's my structure:first,the ground with the PEC

boundary,then a substrate with a patch over it and a via in the middle of substrate,then above all

there's an air box with height of lambda/4.Then I use master/slave boundary for this airbox,with two

variables.phase1 and phase2.At last,use PML boundary at the top of the airbox.The minimum frequency

is 1Ghz,Delta F is 1%,50 steps.For parametricSetup,phase1=0deg,phase2 ranges from 0-180

deg.Everytime after simulation,HFSS9 tells me Write failure in subprocess %3.By the way my computer

has 1G RAM.
I wonder if anybody could help.And thanks very much!If you like,you can also contact with me by

alex_cn6@yahoo.com

pls attach your model, and somebody can help you!

You may try have a look at convergence plot. Right clic in your "solution" tab and look for "Convergence" The click in covergence plot. You shoul see a blue graph and a red line. It doesn't matter if your model does not convergence but if the blue line is quite close to the red one, everythign is ok.

In addtion you may try changing the number of passes (try 20 o 30 passes) ans the acurrency from 0.1 to 0.001. This takes much more time but may make yor model convergence.

Hope this helped!

thanks for all your suggestions.Actually I tried more than 40 steps,but unfortunately all the trials failed.And the attached file is my simulation.Hope someone can give me more suggestions.Thanks very much!

me too!!

I have same problem. It can't converge. also some modes with strong imaginary part. I am not sure what that means? (high dissapation, so actually can't exist?). Also I am not confident how to set the PML layer thickness, and air box length, will it effect the result?
any suggestion highly appreciated!