Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Is a radiation boundary for a simulation necessary?

Status
Not open for further replies.

Heikoo

Newbie level 6
Joined
Nov 6, 2008
Messages
14
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,367
Radiation boundary

hello,

i have following problemm with hfss:

I have a dipol antenna on dielectric subtrat. If i change the distance of the boundary to the model the s-paraemeter diadram will change. Ok. i to distinguish betwen near and far field. What's the rule to assign the size of the radiation box?
Is a radiation boundary for a simulation necessary?
 

Re: Radiation boundary

Yes, u need to use radiation BCs when dealing with antennas. Radiation boundaries must be placed at a distance of at least λ0/4 at the operating frequency (or at the highest frequency of the simulation range) from the radiating surface. That is because of numerical accuracy and volume discretization matters. The minimization of numerical errors (numerical reflections typical of FEM method between the discretized space and the outer non-material non-discretized volume) when u put the radiation boundary depends on the distance from the radiator itself. So I say at least, but higher distances are better. Obviously u must consider an increase in the computational time as u increase volume size. λ0/4 is a good tradeoff.

Radiation BCs are good for a mainly normal incidence at airbox faces but in the case of non-normal incidence a better solution is the use of a Perfectly Matched Layer (PML). HFSS allows the automatic creation of PML boundaries by an user-friedly tool. As for ur case, I'm sure normal radiation boundaries are good enough.

Hope this helps.

I.
 

Re: Radiation boundary

Thanks a lot for your detailed answer. By changing to bigger sizes of the air box for determined sizes the peak of the resonance frequency changes. That means the resonance frequency decreases. For much larger distance to the antenna the resonance frequency is mostly stable and becomes the old value, but why?
 

Re: Radiation boundary

Theoretically speaking, the simulated resonance frequency of the antenna must not depend on the airbox size but it does. The answer is that the simulation is affected by the amount of meshing tethraedra involved and thus by the volume size. Increasing the number of adaptive passes, reducing the convergence limit value, the solution tends to be more stable and closer to reality.

As I said in the last post the distance of the air box is also a value to be chosen properly. When we talk about numerical methods used in simulators like FDTD or FEM (the one used in HFSS) the separation between the simulated region (in a few words where u apply the mesh) and the outer region is of crucial importance. The boundary of the simulation domain is represented by a radiation boundary (which applies the Sommerfield Condition, Far Field Condition hence a of near2far field transformation). The 'mismatch' between the meshed region and the far field region represented by means of a radiation boundary implies numerical errors called 'numerical reflections', which are non-physical reflections at the domain edge due to such a discontinuity. It is proved that this reflections tend to reduce as the distance from the radiating/scattering element and the domain borders gets farther.

I hope now it's clear and u understood why larger distances implies a more stable solution.

I.
 

Re: Radiation boundary

I am understanding the reasons for the differences. But why you suppose to calculate λ/4 from the highest frequency in the simulation and not from the lowest with largest wavelength.
 

Re: Radiation boundary

Yes, u are right. I wrote highest but I meant the lowest, of course, where u have the 'longest wavelength'.
 

    Heikoo

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top