wandering heart
Member level 5
- Joined
- Feb 10, 2006
- Messages
- 93
- Helped
- 5
- Reputation
- 16
- Reaction score
- 3
- Trophy points
- 1,288
- Location
- Stockholm, Sweden
- Activity points
- 1,734
Hi, all~
I want to conduct mode analysis for an antenna. This method is used by some researchers. I searched the internet and got several papers. However not much detail on how to model can be obtained there.
About the model setup, PML seems to be necessary. Other hard boundaries are not reasonable to use for computational domain termination. The is also the choice of several authors.
As for simulation tools, CST, HFSS and COMSOL are accessible to me. CST does not support eigenmode analysis with PML. COMSOL nearly does not perform, with lots of various error messages. HFSS is also seen to be used by some authors. So I'd like to try HFSS.
In my model, I put the antenna at the center of an air box surrounded by PMLs. The model runs well without many problems. But correct results still do not come out due to several issues.
The eigenfrequencies calculated by HFSS are always quite near the specified minimum frequency, no matter what value I input. It seems that HFSS can get many eigenfrequencies everywhere. So I'm quite confused how HFSS determine an eigenfrequency. In my opinion, this phenomenon may be because the eigenmode solved is that for the whole airbox, rather the single antenna. Furthermore, is it possible to change its criterion so that I can find fewer and more significant eigenfrequncies?
Very slow convergence is also bad. This is believed to be due to using lossy materials (PMLs are lossy) in eigenmode analysis. Are there any else boundary conditions suitable to my problem?
Thanks a lot for your patience of reading such a long paragraph!
More thanks goes to people who reply
I want to conduct mode analysis for an antenna. This method is used by some researchers. I searched the internet and got several papers. However not much detail on how to model can be obtained there.
About the model setup, PML seems to be necessary. Other hard boundaries are not reasonable to use for computational domain termination. The is also the choice of several authors.
As for simulation tools, CST, HFSS and COMSOL are accessible to me. CST does not support eigenmode analysis with PML. COMSOL nearly does not perform, with lots of various error messages. HFSS is also seen to be used by some authors. So I'd like to try HFSS.
In my model, I put the antenna at the center of an air box surrounded by PMLs. The model runs well without many problems. But correct results still do not come out due to several issues.
The eigenfrequencies calculated by HFSS are always quite near the specified minimum frequency, no matter what value I input. It seems that HFSS can get many eigenfrequencies everywhere. So I'm quite confused how HFSS determine an eigenfrequency. In my opinion, this phenomenon may be because the eigenmode solved is that for the whole airbox, rather the single antenna. Furthermore, is it possible to change its criterion so that I can find fewer and more significant eigenfrequncies?
Very slow convergence is also bad. This is believed to be due to using lossy materials (PMLs are lossy) in eigenmode analysis. Are there any else boundary conditions suitable to my problem?
Thanks a lot for your patience of reading such a long paragraph!
More thanks goes to people who reply
