Philip,
If an oblique angle of incidence is your game, you'd better not try to
do so using a phase factor amp-function with a gaussian source! Any
single choice of position-dependent phase factor will only succeed in
tilting each of the infinitely many frequency components by a
different angle. However, the prescription for tilting a fixed-
frequency plane wave could be generalized to a huge collection of
super-imposed continuous source plane waves (each with their own
appropriate amplitude functions), since the total field is frequency-
separable. Effectively this is a "digitized" tilted gaussian source.
You'd just better believe your frequency-domain results at these
specific "stepped" frequency channels only. However, note that the
simulation speed suffers as more and more individual sources need to
be initialized.
Now, since a "digitized" source adds together like a Riemann sum, you
should be able to do some integration of the relevant quantities on
paper to find a closed-form amplitude function A(x,t) that will give
you a true tilted gaussian source, or any frequency-domain shape
distribution for that matter. I haven't done this math however, and I
expect it's not pretty.
In most cases, I think you're better off tilting your geometry rather
than tilting the source, you get the same effect. The only problem
then is if you still need to have Bloch-periodic boundary conditions
with respect to the geometry. If that's the case, I guess you're in
trouble?
Regards,
Alex
____________________________________________________________________
Alexander S. McLeod
B.A. Physics and Astrophysics - University of California at Berkeley
Simulation Engineer - Theory Group, Molecular Foundry (LBNL)
Site Lead - Network for Computational Nanotechnology at Berkeley / MIT
asmcl...@lbl.gov 707-853-0716
____________________________________________________________________
On Dec 3, 2009, at 12:45 AM, meep-discuss-requ...@ab-initio.mit.edu
wrote:
Thanks Alex for ur suggestions.
Though I haven't played around with 3D simulations (so far) I didn't
realize that creating a 3D plane wave is so easy with a source plane
where a source line in 2D.
Similarly to 2D, i guess the amp-func property should be expressed
appropriately (exp(2*pi*kx*x)) if the 3D plane wave needs to be
tilted.
But an interesting point from http://article.gmane.org/gmane.comp.science.electromagnetism.meep.general/3058
is that if the plane-wave is a gaussian-src, as in your example,
could we give the amplitude of kx as 2*pi*fcen? So, isn't the most
correct way to create a tilted plane-wave is what is suggested by
that thread, ie. instead of changing phase of the line-source along
the line the line itself is tilted.
Really appreciate your comments.
Thanks again
--
philip
_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
