Assigning symmetry planes in EM simulators

To all the EM fans,

I had a doubt in modelling structures in MWS or HFSS.How will you determine the nature of symmetry ? Meaning when does E symmetry hold and when does H symmetry hold.This is basically an EM problem.How do you determine whether tangetial H or E is zero along a particular plane in a structure say p@tch antenna ?

Thanks for listening

-Arun

my understanding is that before we assign the symmetry plane to perfect E or H plane is depend on what kind of bounday condition at this plane.
If it is along the perfect conductor, we have the pre-knowledge that E field dont have tangential component along the surface. So we use perfect E plane. Same situation to perfect H plane.
In the case of the interface of different dielectric materials, we use perfect H plane, because of H field's continuity nature(permeability of different material are same in most cases).
For exmaple, we analyze the microstrip line, use the center symmetric plane to cut half, we need set this plane to perfect E and perfect H, simulate two times, so we can get the even mode and odd mode.

But in HFSS modeling, we just set this plane to perfect H, then can simulate, I am also confused by this problem. Because microstrip can not support exact TEM wave, that means the E field have components along the trace surface, so how should we set up the boundary condition at this symmetry plane?

Need some comments~

Best Regards,

You gotta have prior knowledge about your model. You should at least have some idea how the electromagnetic field would behave within your model. The best examples for this case are the CPW and rectangular waveguide.

1.CPW
We know that it consists of one signal line and two ground line on left and right side of the signal line. The CPW mode is that the electric field will equally distributed from the signal line to the ground line. The electric fields are are parallel to the symmetry plane. Therefore, perfect-H symmetry can be used. The same can be done to microstrip.

2.Rectangular waveguide.
The fundamental mode is TE10. In this mode the electric fields are drawn as straight lines from bottom to metal to top metal or vice versa (depending on the phase). Taking the horizontal symmetry plane, this plane can be considered as perfect-E plane since the electric fields are perpendicular to the symmetry plane.

Indeed, the CPW and microstrip do not support TEM mode. They are Quasi TEM. This has nothing to do with the symmetry plane, since it is something regarding the propagation direction.

In case of patch antenna, it is not really suggested to have symmetry boundary condition. However, you can alway try simulate it once and see if such symmetry exist.

I think some interesting properties of symmetry you can find in manual of Fullwave. This is FDTD simulator for photonic design.
From my little experience with FDTD and pbg I can mentioned that in order to calculate symmetric or antisymmetric mode it is enough to apply symmetric or antysymmetric boundary conditions with respect to planes.

As far as I understand, you have to consider two things:
1) symmetry of the excitation signal
2) symmetry of the structure

wlcsp said:

You gotta have prior knowledge about your model. You should at least have some idea how the electromagnetic field would behave within your model. The best examples for this case are the CPW and rectangular waveguide.

1.CPW
We know that it consists of one signal line and two ground line on left and right side of the signal line. The CPW mode is that the electric field will equally distributed from the signal line to the ground line. The electric fields are are parallel to the symmetry plane. Therefore, perfect-H symmetry can be used. The same can be done to microstrip.

Click to expand...

The E field in the substrate between trace and ground plane, no problem, perpendicular to trace and ground, so use pefect H, But, the exact mode of microstrip line is not TEM moed, there must be E field component along the surface of trace and ground. If you use perfect H, that means you neglect this part of field, is it?

asdfaaa said:

wlcsp said:

You gotta have prior knowledge about your model. You should at least have some idea how the electromagnetic field would behave within your model. The best examples for this case are the CPW and rectangular waveguide.

1.CPW
We know that it consists of one signal line and two ground line on left and right side of the signal line. The CPW mode is that the electric field will equally distributed from the signal line to the ground line. The electric fields are are parallel to the symmetry plane. Therefore, perfect-H symmetry can be used. The same can be done to microstrip.

Click to expand...

The E field in the substrate between trace and ground plane, no problem, perpendicular to trace and ground, so use pefect H, But, the exact mode of microstrip line is not TEM moed, there must be E field component along the surface of trace and ground. If you use perfect H, that means you neglect this part of field, is it?

Click to expand...

This is the Quasi-TEM approximation for the microstrip right.I think you are safe in using this upto something like 5 GHz beyond which you have to use this model with care.

-Arun

I agree that it is quasi-TEM mode, based on the symmerty nature of this mode, we can use perfect H plane to truncate the computation domain. But, what will you do if we need calcualte the exact mode field of microstripe line(conductor is very lossy metal)? HFSS is a full wave EM solver, in order to use symmetry plane to save somputational cost, how should we assign the bounday condition to the symmetry plane?

Best Regards,