Reflection/transmission not defined due to possible grating lobes

Juan David de la Fuente

Hello there,

I'm trying to develop a study with the RL-GO solver, so I also need to run a 'pre-study' to create the .tr file for the medium. One of the parts in question is made of construction_concrete.

In the 'pre-study' I'm creating a layered structure of 5mm per side with the following properties:

• Thickness: 0.1mm (due to WARNING 2757: A triangle is too thick as compared to the lateral dimensions for application of the skin effect approximation)
• Dielectric material: Construction_concrete_block

As the Example H05. Jerusalem_Cross_FSS specifies, I'm including a two dimensions PBC with 5.2mm per side, the operating freq. is 77GHz and running the solver the following WARNING appears:

WARNING 49005: Reflection/transmission not defined due to possible grating lobes.

 

I reviewed other similar questions here at the forum (Interpretation of the near field) and one of the alternatives is to move the measurement of the transmission/reflection coeff. to 10 wavelengths, lambda=c/f, 3x10^8/77=3.896mm. If I request the transmission/reflection coeff. at:

• X: 4mm
• Y: 0mm
• Z: 0mm

the WARNING still appearing. I think using near field request for 'Calculate only the scattered part of the field' is not the more appropriate alternative since my full model intended to run with RL-GO is in meters. 

 

Should I change something on the study definition? Or this is something I can take as an 'acceptable' WARNING?

Should I change the solver for the full model?

 

Thanks for your support on this,

-David 

 

MartinV_20771

Hello David,

 

The reason for the message about grating lobes is that the size of the unit cell (the PBC) is larger than half a wavelength. A periodic structure larger than half a wavelength has grating lobes. In the next version of Feko we will be adding the capability to compute TR coefficients in such cases as well. For now, better is to make the unit cell smaller if possible.

If your application involves flat concrete without any special periodicity, then you can use RL-GO without a 'pre-study' to obtain T/R coefficients.

 

Best regards,

 

Martin.