Infinite Ground plane

Altair Forum User

Hi,

 

I am trying to apply an infinite ground to my transmitter to simulate my transmitter more realistically and to determine the radiated E-field in its near field region.

 

So i tried using the PEC ground plane, but the far field pattern is giving some strange looking results, and actual transmitter is surrounded by soil/concrete.

 

I tried simulating with the Sommerfeld Integrals but MLFMM is not able to solve with special green's functions.

 

I seem to be only able to simulate with the Reflection coefficient approximation. 

 

Question:

1) Would it be more accurate to simulate it with Sommerfeld Integrals? Any way around this as my structure is rather large.

2) Can I use the infinite planar multi-layer substrate to define an infinite earth ground plane? How is this different from the Sommerfeld integrals?

3) Any idea how accurate the reflection coefficient approximation is compared to Sommerfeld integrals?

 

Thanks in advance.

 

 

JIF

Hi CAW4933,

 

So i tried using the PEC ground plane, but the far field pattern is giving some strange looking results, and actual transmitter is surrounded by soil/concrete. 

What do you mean my strange results. I don't expect any problems with PEC and with a reflection coefficient approximation you should be fine as long as the structure is not too close to the ground.

 

1) Would it be more accurate to simulate it with Sommerfeld Integrals? Any way around this as my structure is rather large. 

They are the exact integrals and would be more accurate and would be required if you want to calculated fields withing the layered structure, but the approximations should be good as long as you don't need measurement inside the layers (below the ground) and for the real ground (dielectric) your structure is not too close to the ground.

 

2) Can I use the infinite planar multi-layer substrate to define an infinite earth ground plane? How is this different from the Sommerfeld integrals?

No, the infinite planar multi-layer substrate is the implementation of the Sommerfeld integrals.

 

3) Any idea how accurate the reflection coefficient approximation is compared to Sommerfeld integrals?

I have not done a study, but I would encourage you do do it. Use the MoM with the reflection coefficient approximation and the Sommerfeld integrals and post your results here. You probably want to do the test with a simplified structure so that the MoM does not require too much resources - the size of the model should not have an effect on the study.