Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Modeling a 3D EM plane wave

Please login with a confirmed email address before reporting spam

As a preliminary exercise to a more complex problem I simply want to model a propagarting electromagnetic plane wave in a block shaped homogeneous medium. But with even such a simple study I have problems.

1. How can I excite the computational domain, i.e. how can I introduce the plane wave into the region of interest? I tried "Port" boundary condition but it just never gives back a normal propagating wave with sinusoidal phase variance. Many times it does not even apply the boundary conditions: where the plane wave should be with a definite phase and amplitude I find a distribution of EM field values. How can it be?

2. What kind of boundary conditions should I use on the lateral sides and at the end of the domain?

Thanks

4 Replies Last Post 2012年9月25日 GMT-4 08:57

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年9月24日 GMT-4 11:19
First of all, make sure you are plotting the right quantity.

1. "Port" BC is what you need. You might have a "distribution of EM field values" if you wave is not perpendicular to the surface.
2. PML for finite domain and Floquet for periodic BC. In some cases, PMC and PEC might work, but try avoiding them.
First of all, make sure you are plotting the right quantity. 1. "Port" BC is what you need. You might have a "distribution of EM field values" if you wave is not perpendicular to the surface. 2. PML for finite domain and Floquet for periodic BC. In some cases, PMC and PEC might work, but try avoiding them.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年9月24日 GMT-4 11:32
If you are trying to simulate scattering of EM wave, you may find solving relative electric field helpful.

Cheers!
If you are trying to simulate scattering of EM wave, you may find solving relative electric field helpful. Cheers!

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年9月25日 GMT-4 08:24
Hi, thanks for the answer.

But my problem is that even when the wave is perpendicular to the surface, I find different values for differet locations on the boundary.

For example, if I set that Ex=1 V/m on a face of a cube, and therefore Hy is automatically set accordingly (propagation is in the z direction), I find nonzero values for Ez and Hx on the port boundary despite that they are prescribed to be zero by the BC. I use PML so no backpropagation from imperfect matching can occur. I just don't understand. Comsol seems to be incapable of modeling a simple propagating plane wave in a medium.
Hi, thanks for the answer. But my problem is that even when the wave is perpendicular to the surface, I find different values for differet locations on the boundary. For example, if I set that Ex=1 V/m on a face of a cube, and therefore Hy is automatically set accordingly (propagation is in the z direction), I find nonzero values for Ez and Hx on the port boundary despite that they are prescribed to be zero by the BC. I use PML so no backpropagation from imperfect matching can occur. I just don't understand. Comsol seems to be incapable of modeling a simple propagating plane wave in a medium.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年9月25日 GMT-4 08:57

Hi, thanks for the answer.

But my problem is that even when the wave is perpendicular to the surface, I find different values for differet locations on the boundary.

For example, if I set that Ex=1 V/m on a face of a cube, and therefore Hy is automatically set accordingly (propagation is in the z direction), I find nonzero values for Ez and Hx on the port boundary despite that they are prescribed to be zero by the BC. I use PML so no backpropagation from imperfect matching can occur. I just don't understand. Comsol seems to be incapable of modeling a simple propagating plane wave in a medium.



Hi Daniel,

Maybe you miss one thing. When using the port boundary condition inside your domain you have to ensure the mesh on either side of the port is the same. Use copy mesh for this.

Try first perhaps with perfect conductor on the side. Say that the incoming wave has only electric field in the Y polarization and your wave propagates in the z direction. Put PEC boundary condition in the Y sides of your square domain. and on the X direction put PMC (perfect magnetic conductor). On the top and bottom your port boundary condition. If for some reason this does not work, try first solving for scattered field. You would be required to input an expression of the incoming field directly if you choose this mode (in the electromagnetic waves node of your model builder). For the scattered field calculation just put PML on the top and bottom of your domain instead of the ports. Solve with direct methods to be sure nothing funny happens for the first try.

Hope this helps.

Aimi

Best regards,


Aimi
[QUOTE] Hi, thanks for the answer. But my problem is that even when the wave is perpendicular to the surface, I find different values for differet locations on the boundary. For example, if I set that Ex=1 V/m on a face of a cube, and therefore Hy is automatically set accordingly (propagation is in the z direction), I find nonzero values for Ez and Hx on the port boundary despite that they are prescribed to be zero by the BC. I use PML so no backpropagation from imperfect matching can occur. I just don't understand. Comsol seems to be incapable of modeling a simple propagating plane wave in a medium. [/QUOTE] Hi Daniel, Maybe you miss one thing. When using the port boundary condition inside your domain you have to ensure the mesh on either side of the port is the same. Use copy mesh for this. Try first perhaps with perfect conductor on the side. Say that the incoming wave has only electric field in the Y polarization and your wave propagates in the z direction. Put PEC boundary condition in the Y sides of your square domain. and on the X direction put PMC (perfect magnetic conductor). On the top and bottom your port boundary condition. If for some reason this does not work, try first solving for scattered field. You would be required to input an expression of the incoming field directly if you choose this mode (in the electromagnetic waves node of your model builder). For the scattered field calculation just put PML on the top and bottom of your domain instead of the ports. Solve with direct methods to be sure nothing funny happens for the first try. Hope this helps. Aimi Best regards, Aimi

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.