Hello, I have a question concerning the near surface field values in FEKO. I understand that FEKO is computing surface current and make them radiates with Green function to have field in space. When I ask the field value near a surface, I can see the field growing as 1/r. This is expected since this is the formula for a current radiating a magnetic field. However, if I keep asking values closer and closer to the surface, I can get the maximum value near the surface higher and higher. I understand that I probably begin to see numerical aberrations rather than the real magnetic field value. My question is : * What is the confidence I can have about my value as I get closer to the surface ? I understand that I also have to mesh finer and finer as I get closer to the surface. But in the end, I expect my magnetic field to converge toward a value. Can I have this value ? * I understand that FEKO computes surface current, which means I maybe could get my field as high as I want ? Getting closer and closer though, my maximum value is kind of growing very slowly, so the opposite behaviour of a 1/r formula. So I guess it is still possible somehow to get the real surface field value ? Two graph of the field as I get closer to the surface : Thanks a lot for reading, Pierre

Near surface field values with FEKO

Hello,

I have a question concerning the near surface field values in FEKO.

I understand that FEKO is computing surface current and make them radiates with Green function to have field in space.

When I ask the field value near a surface, I can see the field growing as 1/r. This is expected since this is the formula for a current radiating a magnetic field.

However, if I keep asking values closer and closer to the surface, I can get the maximum value near the surface higher and higher. I understand that I probably begin to see numerical aberrations rather than the real magnetic field value.

My question is :

What is the confidence I can have about my value as I get closer to the surface ? I understand that I also have to mesh finer and finer as I get closer to the surface. But in the end, I expect my magnetic field to converge toward a value. Can I have this value ?
I understand that FEKO computes surface current, which means I maybe could get my field as high as I want ? Getting closer and closer though, my maximum value is kind of growing very slowly, so the opposite behaviour of a 1/r formula. So I guess it is still possible somehow to get the real surface field value ?

Two graph of the field as I get closer to the surface :

Thanks a lot for reading,

Pierre

Find more posts tagged with

Feko

Accepted answers

All comments

Mahan Rudd_22044

Hi @degleve_jarmlami ,

There is no single value that can be declared for the fields to converge as it depends on lots of factors. You can do the mesh convergence check (refine the mesh by 10% at each step) and see how it affects the results. if there is no change in the results then the results are converged (mesh convergence check). There are some features such as error estimation that can be used to assist with the manual refinement. The other option (automated) is to use AMRFEKO.

The fields on the surface are NOT going to be as high as you want. You are right about the the fields getting higher as you get close to the surface (1/R behavior) but it will reach to a FINITE value on the surface (R=0). You noted that Feko calculates the surface and uses the Green's function to calculate the scattered fields:

Escat(r)=∫SJ(r′)G(r,r′)dS′

G ~ 1/R

at R=0 , you will have ∫1/R dS′ and dS′∼RdRdϕ so the singularity cancels at R=0 and the fields will be finite. This is consistent with the Maxwell's equations boundary conditions.

Regards

Mahan

degleve_jarmlami

Hello,

Thanks for answering:) understand. Indeed with finer mesh and closer field points, the value seems to converge nonetheless !

Pierre