On 06/22/2017 01:39 AM, mayawei wrote:
But if there are more than one ω(n is not one),which one it
returns? That means how can I understand (define f (harminv-freq-re
(harminv-freq-re harminv-results) will return an error if there is more
than 1 frequency. To return the nth element in the list, you would do:
(harminv-freq-re (list-ref harminv-results (- n 1))
Also, as explained in the Reference page, to obtain a list of all
frequencies, you would do:
(map harminv-freq-re harminv-results)
In general, if you want to find out what each Scheme/libctl function is
doing, check the definitions in the source file: meep/libctl/meep.scm.in.
So there are two run-sources+ (run-sources+ Th (after-sources (harminv
Ez (vector3 0) fcen df)) and run-sources+ T (dft-ldos f 0 1))
corresponding to
two formulas.From the first run-sources+,we get f in (dft-ldos f 0
1).And if I just want to do (dft-ldos f 0 1)) ,how can I get f without
the first run-source+?If there more >
than one f(or ω),how can I deal with?
If harminv finds more than 1 frequency, you can include multiple
(dft-ldos ...) run functions to compute the LDOS at each frequency:
(run-sources+ T (dft-ldos f1 0 1) (dft-ldos f2 0 1))
However, note that the quality factor and hence the loss rate (via the
"T" parameter) for each mode may be quite different. Thus, it may be
more accurate to have separate runs, one for each frequency.
Third,in the formula (2) ,if the material is metal and its epsilon
is complex(Lorentz-Drude model),can this formula apply and how can I
deal with?
The LDOS formula of eq. 2, which is eq. 4.22 in our book chapter:
http://arxiv.org/abs/1301.5366, was derived for a lossless system (i.e.,
purely real permittivity). It does not apply to lossy metals where the
finite loss modifies the calculation of the work done by the electric
field on the electric current.
Forth,in the formula (1) and (2), ω is (harminv-freq-re result) but
not source frequency fcen ,why not (harminv-freq result) ?
The frequency of the closed metal box can be computed analytically which
we use as the center frequency of the Gaussian pulse. However, because
the notch perturbs the system, we need to do an initial calculation to
determine the mode frequency which has shifted. It is at this shifted
frequency where we compute the LDOS at the center of the box. We must
use the real part of this frequency to pass to the (dft-ldos ...)
routine since this involves calculation of the discrete Fourier
transform (DFT) of the fields which is based on real frequencies.
_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
