Dear Meep Users,
I'm currently using meep to find the modes of a plasmonic cavity
(metal-dielectric-metal). I use a drude model for silver, which I've coded
into the control file using the procedure described here:
http://ab-initio.mit.edu/wiki/index.php/Meep_Tutorial/Material_dispersion(and
attached below). The problem occurs when I use harminv to find the
resonant modes. I use a broad gaussian pulse (centered around my expected
resonance frequency) and use harminv in a fashion similar to the
ring-resonator tutorial (I only run my simulation for 100 time units after
the sources as I expect low Q-factors). I find that the mode frequencies and
quality factors are very sensitive to resolution. Unfortunately, I am having
trouble finding convergence in my solution. At a very high resolution of 120
or 1.2 pixels per nanometer (my unit of length is 100 nm), I obtain
reasonable results. Yet, when I increase the resolution to 140, Harminv
finds modes with low negative quality factors and shifted frequencies.
I would appreciate any insight on increasing the numerical stability of this
simulation. Normally I would simply increase resolution, but that doesn't
seem to be helping. Thanks.
Ag Drude Model:
;Silver Drude Model Parameters in SI
(define-param wp 1.4433e+16) ;This is in Rad/s
(define-param f_relax 1e+14) ;This is in Rad/s
(define-param e_inf 5)
;Drude Model Parameters for meep format and in meep units
(define unitw (/ (* 2 pi c) unitL))
(define gammaMeep (/ f_relax unitw)) ; gamma in base unit
(define wpMeep (/ wp unitw)) ; resonant plasma frequency in
base unit
(define de (* wpMeep wpMeep 1e+40)) ;in meep de = wpMeep^2/omega^2
(make dielectric (epsilon e_inf)
(polarizations
(make polarizability
(omega 1e-20) (gamma gammaMeep) (sigma de))
)))
--
Carlos Octavio Aspetti
Graduate Student
University of Pennsylvania
Department of Materials Science and Engineering
_______________________________________________
meep-discuss mailing list
meep-discuss@ab-initio.mit.edu
http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
