Anisotropic slab periodic boundary condition
I used PBC to define an infinitely extended anisotropic slab. I highly doubt the near field results predicted by feko, as the total field is approximately equal to the incident field
<?xml version="1.0" encoding="UTF-8"?>
I have attached the CADFEKO model, can you let me know if the PBC are defined correctly.
Answers
-
It seems the properties of the inside anisotropic cuboid was lost.
It seems when you defined the small and larger cuboids, one of them was set to the SEP method and the other to FEM.
After the Union the properties of the inside cuboid was merged with outer (air).
(Sidenote: You don't have to make the unit cell so large. A length or width of about half a wavelength should be ok, but if you make it too small the interpolation will start to break down.)
See attached image of the volume meshes.
Incorrect mesh:
Correct mesh:
<?xml version="1.0" encoding="UTF-8"?>
0 -
Hi Mel,
Can you please attach your model with the correct mesh.
Thanks
0 -
As you mentioned I reduced the size of the slab and checked if the mesh are defined correctly for different regions. But the total near field result looks still not right
<?xml version="1.0" encoding="UTF-8"?>
0 -
I found out that when I don't use periodic boundary condition the mesh looks correct as below
<?xml version="1.0" encoding="UTF-8"?>
but when I specified periodic boundary condition and re-meshed the model then the anisotropic region gets defined as air.
<?xml version="1.0" encoding="UTF-8"?>
Which causes the total scattered field to equal to the incident field. Is this a bug in FEKO when using PBC?
0 -
When I ran the attached model with the correct mesh the feko solver reports the following error message
ERROR 50228: The boundary mesh elements of the unit cell are not consistent with the periodic boundary condition
Can you please provide a solution for this issue
0 -
Please install Feko 2019 and then run the updater to get version 2019.3.
Attached my model with a smaller extent just for testing purposes. Neither the meshing nor Error 50228 occurs here.
0
