How to solve ERROR 47180: MoM segments cannot be in contact with RL-GO triangles

Otta8

I have a mesh of an airplane on witch I have construct a monopole lambda/4, wire radius lambda/20 at 1.5GHz. My request is refering to far field. I want to solve it with RL-GO but it returns the following error:

"ERROR 47180: MoM segments cannot be in contact with RL-GO triangles"

First question: when I construct the line, I can not choose the kind of solution that I want. Is it because Feko implement only MoM in this case?

Second question: I read the documentation about how Feko implements the RL-GO solution and it says that "Feko integrates the RL-GO method with the current-based MoM, by launching rays from each radiating MoM element". So it seems to me that RL-GO method in Feko does not create triangles. Why then does the description of the error speak of triangles?

Third question: I tried to use as a solution the PO - full ray - tracing for large elements and in this case it works. Why solving with PO it works while with RL-GO doesn't?

If someone can tell me how to solve this problem, it would be great.

Thank you.

Torben Voigt

Hi @Otta8 ,

RL-GO can only be applied to faces, not wires. And yes, wires are always solved with MoM. 

PO and RL-GO are bothe "high frequency methods", feasible for models with a minimum size of ~20 lambda. Still the methods are very different:

PO & LE-PO – Physical Optics & Large Element Physical Optics

• First the MoM area is calculated (in your case the wire). Then Feko checks whether a PO area is "illuminated" by the magnetic field, and then surface currents are calculated accordingly. In case of coupled MoM / PO, the back-coupling on the MoM area is then calculated and the MoM region is recalculated accordingly. The procedure is repeated until the reaction of the PO to the MoM region is fully taken into account.
• With LE-PO you can use triangles of any size (as long as there is no near field request and source far away), because the current along the edges does not have to be constant.

RL-GO – Ray-Launching Geometrical Optics

• Strongly simplified description: From each source (e.g. each edge of a MoM area) beams are fired. Where they hit a RL-GO surface, a substitute source is placed, which in turn fires beams again (reflection and transmission in dielectrics).
• As you can imagine, there must be a certain distance between MoM elements and RL-GO faces, so that beams can be launched. MoM and RL-GO must not touch.

If your model is not too large, I would recommend to try MLFMM first before switching to asymptotic methods.

I also attached a very brief description of the solvers in Feko (in case you didn't ever have a Feko training).

 

Best regards,
Torben Voigt