Elmer Discussion Forum

Specifically, I would like to find the (far-field) radiation pattern of a microstrip antenna – you know, the kind of antenna that you etch into the copper of a printed circuit board (the patch antenna being the most well known example).

In case Elmer is a good match for this, is there an example of someone doing the same kind of simulation, whose lead I could follow?

Hi

Depends. I assume radiation patters require the solution of electromagnetic wave propagation (described by a vector-Helmholtz equation). That we have in both harmonic and transient mode using modern edge element formulation.

However, I also understand that ideal antenna simulation would have features allowing the definition of the problem in terms of the antenna designer. That we don't have.

So the suitability of Elmer propably depends on your willingness to bridge the gap from generic EM-wave solver to something useful in your field. This certainly would be an interesting application field.

-Peter

raback wrote: ↑19 Nov 2018, 20:57 I assume radiation patters require the solution of electromagnetic wave propagation (described by a vector-Helmholtz equation). That we have in both harmonic and transient mode using modern edge element formulation.

That sounds good. Where can I learn about this edge element formulation?

raback wrote: ↑19 Nov 2018, 20:57 However, I also understand that ideal antenna simulation would have features allowing the definition of the problem in terms of the antenna designer. That we don't have.

I think all that would be needed is a way to specify a voltage (sinusoidal in time, for example) between two feed points (or surfaces) on the simulated structure. Alternatively, a current source could be used instead of a voltage source, if that makes it easier. It would be good to be able to give the current or voltage source an output impedance too.

Hi, I think the BCs you ask are pretty standard stuff. Should be no problem. -Peter

raback wrote: ↑29 Nov 2018, 16:42 Hi, I think the BCs you ask are pretty standard stuff. Should be no problem.

Cool! Any pointers for where I might learn how to set up such a simulation (of a printed circuit board, i.e. of two, flat copper structures with a slab of dielectric in between), and such boundary conditions? Maybe someone has done something that overlaps with what I'm trying to do, whose work I could learn from. (I haven't used Elmer before.)

Hi

There is not really much.

The equation is presented as Ch. 18 in Elmer Models Manual.

There are only a few consistency tests among the hundreds: which you could try to run to see the light.

I would advice to start very simple and add complexity gradually.

-Peter

Hi Jolson and Peter,

I am also interested in using Elmer for patch antenna simulation and potentially compare results to Ansys HFSS.

Were you successful in setting up an antenna problem and are you aware of such comparisons - my search was not successful so far.

Your help is greatly appreciated.

Thank you!
Andreas

andreas.hoffie wrote: ↑10 Mar 2021, 06:01 I am also interested in using Elmer for patch antenna simulation and potentially compare results to Ansys HFSS.

Were you successful in setting up an antenna problem and are you aware of such comparisons - my search was not successful so far.

In the end, I did not go through with the simulations, so I have no insights to offer.

Hi,

Googling a little it seems Ansys HFSS uses several different numerical techniques. Elmer would be limited to quite a low wavenumber regime since it only has FEM available. Augmenting the code to include MoM or other methods is not planned.

-Peter