How to determine the size of an air box in HFSS?

[jimmyguo2005]

hey, guys, i m a newbie in antenna design. i have got a problem, is there anybody know how to determine the size of the air box? does the size affect the result?

thx!!

 

[pepino]

Re: HFSS AIR BOX PROBLEM

The best is if the walls are on distance greater than λ/4 from the model. But if they touch the model (or in close proximity) there also can be acheved good results in v.9.0- v.9.2 simultor vresions. Version 8 is more sensitive to the size of the box and ports.

 

[kash787]

Re: HFSS AIR BOX PROBLEM

Hi buddy,

The air box walls have to be away at least a distance of λ/4. It will give you good results. Howvere, I wouldnt really use that. the best way to do thid and save computation time is to use something called a PML " A perfectly matched layer". What this boundary does is that it absorbs all radiation incident on it so in theory this is a perfect absorbing layer. You can assign this boundary on air box which is much closer to your antenna thereby saving a lot of computational time. Hope this helps.

If u want an example send me pm
Kash.

 

[elsa]

Re: HFSS AIR BOX PROBLEM

According to my experience ,using PML can get better results,but it'll take much more time than using radiation boundary.

 

[terry andy]

Re: HFSS AIR BOX PROBLEM

if u want to get the farfield-results the air should be defined as radiation boundery!

 

[fishyxl]

HFSS AIR BOX PROBLEM

I agree with above suggestions that PML boundary conditions give better results. But it takes less time, in my opinion!!!!!!!!!!!!

 

[adel_48]

HFSS AIR BOX PROBLEM

Hi,

the absorbing boundary condition should give good results as long as the direction of propagation is perpendicular or with small incidence angle to it (which is why it should be placed at a distance from the geometry)

the PML gives good results for any incidence angle, that is why it needn'e be placed far from the geometry. But it has a disadvantage which is it is unlike the absorbing boundary, the field must be solved in it, thus it requires extra time if it is placed at the same distance as an absorbing boundary.

BR
Adel_48

 

[victorcl]

HFSS AIR BOX PROBLEM

PML may be is not so good ,it neglects the coupling
between antennas

 

[cyai]

Re: HFSS AIR BOX PROBLEM

In typically, we set the air box size to be near quarter-wavelength, then use the funcation, PML, optimize the better box size. But remember, the solution frequency shall be near the wanted frequency.

 

[aboozar.hamidipoor]

Re: HFSS AIR BOX PROBLEM

hi:

the best result is taken when you draw the box with more than 3*λ or 4*λ length in the main lobe direction and no length in directions with no propagation . for example in recangular microstrip patch antenna you should draw a box with 3*λ or 4*λ in the perpendicular direction to patch plane and no length more than substrate dimentions in the patch plane direction .

hope to help you.

 

[rautio]

Re: HFSS AIR BOX PROBLEM

If your antenna is planar, you should be using an unshielded MoM tool, no worry about where the box is, and you don't need to mesh the volume, MoM tools just mesh the surface of the metal. I would recommend Agilent Momentum, it is not the fastest tool, but from results I have seen, it is the most accurate unshielded tool. (I work for Sonnet, a shielded MoM tool vendor and occasional competitor with Momentum.)

If your antenna is not planar, then you must consider the comments of the others above carefully, but I would add one more thing. When you have the box big enough, its exact size will no longer be important. If you have to be sure you have a good result, then do an analysis with what you think is the best box size. Then do a second analysis with another box size. If the change in the answer is equal to or larger than the variation you can tolerate in your design, then you must make the box even bigger. This is the approach we recommend when designing planar antennas using our shielded MoM tool, works quite well. Should work for volumen meshers too. Good luck.