微波EDA网,见证研发工程师的成长!
首页 > 研发问答 > 微波和射频技术 > 电磁仿真讨论 > HFSS Analysis Issues

HFSS Analysis Issues

时间:03-29 整理:3721RD 点击:
When my Simulation is done on HFSS its give me result but when i checked the convergence tab in the Soultion data there appears that not converged what does it means are my results are acceptable or not?

Not converged => don't trust results

but i dony understand what the issue behind that non convergence?

It seems that you have not configured enough mesh refinement steps.
I use another FEM tool, so I don't know the exact names used in HFSS dialogs.

Hi bukero,

Convergence in HFSS is measured in a number of ways, but usually it's measured by scattering parameters.

Specifically, convergence is the evaluated based on the largest magnitude of change of each entry in the simulation's scattering matrix ("delta S"), between each mesh refinement step. The idea is that if the mesh is appropriately refined, increasing the meshing further shouldn't change the results in any significant manner. So if the mesh is being refined, and the scattering parameters are changing by a large degree, then the mesh must not be sufficiently accurate.

So is it possible to deal it by changing the value of Delta S?

If you increase delta S, you get the same bad data, just without a warning.
As I already mentioned in #4, you need more iterations, so that HFSS can refine the mesh until it reaches convergence with small delta S.

should solution frequency responsible to, or play any role in convergence? on what basis the value of solution frequencybe assigned or calculated?

Yes, frequency plays a large role in convergence.

Firstly, typically the initial mesh is created based on the wavelength in each medium.

Secondly, adaptive mesh processes generally will operate on the field profiles at the solution frequency. So if your structure supports several modes (say, for example it is a resonant cavity), then the mode(s) which are present at the solution frequency will dictate what the final mesh looks like -- whereas other modes may be present at higher or lower frequencies.

You want the mesh to adapt itself to your modes of interest -- so keep the solution frequency close by any large resonances, and if you can only choose one generally choose it to be slightly higher that any dominant resonance, such that the mesh is generally finer.

Good Luck!

Copyright © 2017-2020 微波EDA网 版权所有

网站地图

Top