Two objects scattering issue

Teng Chang

Hi all,

For this case, the option in advanced setting, “Calculate only the scattered part of the field" has been enabled. I try to extract the near field information reflected by the sphere only (excluding scattered by the square conductive plate). Is this setting correct? I’m afraid of this result including the scattered part of the field induced by the square conductive plate. If this setting is not correct, how could I extract the scattered part of the field reflected by the sphere only? 

One more question, when I change the radius of the spherical near field from 0.3 m to 0.1 m, the results will be quite different, how to fix the radius to a proper value?

Mel

To get the scattered field from the sphere only, use the scope tab.

I am not sure what you mean by the results change when you change the radius of the near field request. Please be more specific.

Teng Chang

Let skip the second question first, another question is, how can I ensure the scattered field from the sphere include the second path only? (avoid receiving scattered filed induced by the equivalent source like the first path)

Mel

In this case you must use:

This option ignores the radiated contribution from impressed sources (Hertzian dipoles, radiation pattern sources) yielding only the scattered fields.

Teng Chang

Dear Mel,

Thanks for your suggestions, so in this case, should I not use near field as request, even if scope is selected to sphere surface, and ''calculate only the scattered part of the field'' is enabling?

I also do a test as attached, please check.

the scenario for this test:

In order to avoid the coupling effect between two antennas (TX/RX, same pattern same polarization), this scenario is divided into two simulations.

For the first .cfx file (Sphere_Proposed_01.cfx), I try to get the scattered field on the sphere, then these scattered field information will be the source of second .cfx file (RX_Power_FF), and the power scattered from the square conductor will be received by RX antenna.

So I try two methods, one is by by using far field Source and near field source.

And these two results seem like almost the same...

Is the setup correct?

I also check the beam shape (radiated by scattered part of the sphere), the results seem a little bit different by using these two methods (by using far field Source and near field source), could you please check it (you can see the file named "RX_Power_BeamCheck.pfs"), which one is more accurate?

by using far field source, the gain pattern is:

 

by using near field source, the gain pattern is:

 

Mel

In general the near field source is more accurate. It is sampled at exact locations.

But the phase of the far field source points will be referenced to the origin specified. 

 

A better method would be to use the solution coefficient source (Feko 2021.1 and later)

First solve the model with the sphere and plate and far field source (any source) and request model decomposition and specify the sphere as the label.

Feko will "save" the currents on the sphere to be used as a source in the next simulation.

In the next simulation remove the source and add the far field receiving antenna. Remove the sphere and in place of the sphere put a solution coefficient source using the *.sol file obtained in the previous simulation:

This method will be the most accurate.