hfss eigenvalue

Hi, guys.

I have a question about HFSS eigen-analysis convergence.

For my analyzed structure, since I used Master-Slave boundary condition, it seems the frequency convergence is pretty poor.

Some time it reaches theoretical value in one pass but divergence to a weired number in the next one....

So are there any good methods to improve the eigen-frequency convergence in HFSS ???

Thanks in advance

I think you need to specify a mesh seeds near the points the electric field varies fast around. This is the only method I know to improve the convergence. I don't think that master slave boundary play a rule in the convergence speed. But, may be you need to check if you set them up correctly

I am experiencing a similar problem with HFSS. I set up the simulation such that at least 5 runs converge. After a while convergence is achieved. Then in the subsequent 3-4 runs, the convergence is achieved. But suddenly at the 4th or 5th run there is an error of %98. Out of 36 runs, none of the 4 runs converged to the same value after one another. Quite confusing. Mayber there is a bug? Or the solver is very poor?

After several trials and error, I found that if I increase the minimum frequency, the convergence can be improved. But I don't know why, maybe at lower frequency, the matrix generate too many pseudo-eigen value???

I noticed the same problem. It seems that it is very difficult to identify the first eigenvalue for the solver if the first eigenvalue is close to 0. I did not experience this problem with CST MWS. By using CST MWS, I was able to calculate the full band diagram of a 2D photonic crystal. The eigenvalue at k=0 is 0 for such a 2D photonic crystal. But I could not produce such a band diagram by using HFSS

Hi, I am too trying to reproduce the bandgap result form Uniplanar
compact electromagnetic baqndgap (UC-EBG). I have problem to set up proper
airbox and PML.
I know that 3 PML on top of airbox is required. Each Layer has a
diffrent material properties. also, I defined dummy airbox which will
overlap the whole structure, i.e., from the ground plane level up to the
top of PML layers. so, I selectmore than tewnty adaptive passes in order
to achieve convergence.
but, I can't reproduce the bandgap result from this paper.
[1]. Z. Iluz, R. Shavitand Reuven Bauer, "Microstrip Antenna Phased
Array with Electromagnetic Bandgap Substrate," IEEE Transaction on
Antenna and Propagation, Vol. 52, No. 6. June 2004.

also, I have another question. How can i defined TE mode or TM mode
in Eigenmode solution.


Would it be possible to download from somewhere an example of
HFSS/Optimetrics projects for the computation of the dispersion diagrams of 2D structures, such as PBGs.

thanks