Author |
Message |
Patrick Temple (ptemple)
New member Username: ptemple
Post Number: 1 Registered: 09-2003
| Posted on Tuesday, September 23, 2003 - 03:34 am: | |
I am trying to find the electromagnetic resonant modes for a hollow cavity cylinder. I have started by using the standard wave equation for Ez but am having trouble determining the boundary condition for the r=0 axis of symmetry. Should it be Natural(Ez) = 0 for the r=0 line? Thanks for any help, Patrick |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 41 Registered: 06-2003
| Posted on Tuesday, September 23, 2003 - 02:29 pm: | |
If your equation is, as I infer, something like div(grad(Ez))+lambda*Ez=0, then the Natural BC specifies the outward normal component of grad(Ez). At the rotation axis, you want this to be zero, so NATURAL(Ez)=0 is the correct boundary condition. |
Patrick Temple (ptemple)
New member Username: ptemple
Post Number: 2 Registered: 09-2003
| Posted on Tuesday, September 30, 2003 - 05:46 pm: | |
Thanks for your quick reply. That worked for finding the resonant modes for a smooth metal cavity. I am now trying to find the eigenvalues for an axially symmetric cavity which has angled exterior walls. This means that Ez will be non-zero at these walls and the E vector must still be normal to the exterior walls. Is there a good way to set this boundary condition which ties in a whole new variable Er? Am I making this problem more complicated than it has to be? Thanks again, Patrick |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 42 Registered: 06-2003
| Posted on Wednesday, October 01, 2003 - 04:50 pm: | |
Since you now have two variables, Er and Ez, you need two boundary conditions. One of them could be value(Er)=(nr/nz)*Ez, to force E normal to the wall. Is there another condition you can impose on Ez or its derivative? |