Scattering boundary conditons for non-normal incident (using weak-contribution)

Topics: 4.2a, RF Module

Thread index  |  Previous thread  |  Next thread  |  Start a new discussion

RSS FeedRSS feed   |   Email notificationsTurn on email notifications   |   3 Replies   Last post: January 23, 2013 5:28pm UTC
Felix

Felix

January 22, 2013 7:22pm UTC

Scattering boundary conditons for non-normal incident (using weak-contribution)

Hi all,

I'm wondering, if it possible to use the scattering boundary conditions in the RF module to make a boundary transparent for a plane wave with a known wave-vector (which is different from the wave vector of the incident field at this boundary).

To give a simple example, if I, for example, want to simulate the reflections of a plane wave (with a wave vector given by k=(kx,0,kz) and a polarisation of E=(1,0,-kx/kz)) by a metallic half space (surface given by z=0), I know that the wave vector of the reflected field will be k=(kx,0,-kz). Is there any possible to make this simple simulation using scattering boundary conditions (without using PML) and solving for the full field? (I don't want to solve for the scattered field or use PML, since it would be inconvenient for the simulations I'm interested in. Obviously my final goal is not to calculate the reflection of a plan wave at a metal surface.)

Thanks a lot in advance for any advises,
Felix

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Alexander Kuznetsov

Alexander Kuznetsov

January 22, 2013 9:55pm UTC in response to Felix

Re: Scattering boundary conditons for non-normal incident

Do you want to use the same BC for launching plane waves as well as for absorbing the reflected wave with different k? You should use port BC for this.

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Felix

Felix

January 23, 2013 2:38pm UTC in response to Alexander Kuznetsov

Re: Scattering boundary conditons for non-normal incident

Hi Alexander,

Thanks a lot for your quick and friendly reply. Unfortunately, I'm still struggling with the problem. Probably I should explain in more detail what I try to do. This is also the reason while I changed the subject of the thread slightly.

I set up a model adapted to the problem I'm actually interested in by using the weak-form interface of Comsol. Now I want to use one boundary of the problem to excite a plane wave within the simulation area (I make 2d simulations), but this boundary should be also transparent for the reflected wave. Using the notation of my first post the weak contribution is given by the cross product of H and n (the normal vector). Assuming that the magnetic(H) and electric(Ex,Ez) field at the boundary are given by an unknown reflected (Hr,Erx,Erz) and a know incident (Hi,Eix,Eiz) contribution, one ends up with

H=1/ (mu0*omega)*(-kz*Ex-kx*Ez+2*kz*Eix)

But using this weak contribution I only get reasonable results for kx=0, in the other cases the reflected wave is reflected again at the boundary of the simulation area resulting in an interference pattern (see the attached image).

I would be very grateful for any advises,
Felix

Attachments:   Field_pattern.jpg  

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Alexander Kuznetsov

Alexander Kuznetsov

January 23, 2013 5:28pm UTC in response to Felix

Re: Scattering boundary conditons for non-normal incident

I'm not sure how to do it using the weak form. Can you avoid using it? Then port BC would work just fine. If you still want to have weak-form, then check the equations in port BC and try to do something similar.

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse


Rules and guidelines