Hello, I am trying to simulate a thin plate whose thickness has to go smaller and smaller. I have a problem because FEKO keeps meshing the edge of my structure with thin triangles : I understand that since there is a lateral face of small thickness, FEKO tries to keep the ratio of my triangles reasonable. However, I really don't need precision at the edge of my structure. * Would someone has a solution to get a thin plate without having to mesh a lot the edges ? → Right now my structure is a rectangle, which I sweep by my thickness value. I tried defining it rather by simply a rectangle and then associate to it the copper with the thickness equal to my plate thickness. However, it really does not gives the same results so I guess this is not a good way to proceed ? * Also, is it that important to keep edge triangle size ratio small ? Thanks a lot, Pierre

Edge meshing get too small

Hello,

I am trying to simulate a thin plate whose thickness has to go smaller and smaller. I have a problem because FEKO keeps meshing the edge of my structure with thin triangles :

I understand that since there is a lateral face of small thickness, FEKO tries to keep the ratio of my triangles reasonable. However, I really don't need precision at the edge of my structure.

Would someone has a solution to get a thin plate without having to mesh a lot the edges ?

→ Right now my structure is a rectangle, which I sweep by my thickness value. I tried defining it rather by simply a rectangle and then associate to it the copper with the thickness equal to my plate thickness. However, it really does not gives the same results so I guess this is not a good way to proceed ?

Also, is it that important to keep edge triangle size ratio small ?

Thanks a lot,

Pierre

Find more posts tagged with

Feko

Accepted answers

All comments

Jaehoon

Hello Pierre,

As I said in the different question, please share the Feko model here..

Let me check it.

Best regards,

Jaehoon

degleve_jarmlami

Hello,

Sorry. Here it is !

test_flat_th.cfx

degleve_jarmlami

And as I keep reducing the thickness parameter (th) the side keeps meshing a lot !

Jaehoon

Hello Pierre,

1.Because of the thickness of the cuboid (Sweep), it is impossible to control a lot of meshes automatically generated, so I cannot reduce the number of the meshes.

2. I know your goal of this simulation: to compare between 2D plate and 3D plate which was already mentioned in the previous ticket. Let's do this comparison here.

I reduced your model (test_flat_th.cfx) to make two models: Plate_2D.cfx and Plate_3D.cfx (attached),

Plate_3D.zip

and made some comparison as shown below. (I don't think two results are too different each other, if we consider 3D is thicker than 2D geometrically. So Plate_2D is reasonable, as MoM uses 2D plate, not 3D plate (metallic face).)

Best regards,

Jaehoon

degleve_jarmlami

Hello,

Thanks a lot for the tests.

My different behaviour is spotted more on the side of these kind of object.

Here are two simulations at 2.87GHz of a simple rod, of length one wavelength.

If you observe the magnetic field on the plane of the antenna ("side" monitor in my simulation), we see the difference (see files and pictures)

I feel this is because how current spread here. Like for a plate the thickness is used to compute how it will react but the current are still concentrated on a plate, while for a volumetric structure current are spread on the thickness so I guess in the near field differences should be expected ?

Thanks,

VOL :

FLAT :

plate_vol.rar

Jaehoon

I agreed on your explanation about the difference.

Best regards,

Jaehoon