On May 29, 2006, at 4:28 AM, K. Choi wrote:


Description about material dispersion in meep article has,

To implement a Drude model e(w), in which ωn is zero in the
denominator, we just set ω_n to be a very small number (e.g. 1e-20)
and make delta epsilon large to compensate in the numerator.

I don't understand this sentance. If I calculate a metal which has
plasma frequency Wp = 1.1368e+16 and relaxation rate r = 9.7269e+13,
how set this parameters into meep?

take a look at http://ab-initio.mit.edu/wiki/index.php/Meep_Tutorial/ Material_dispersion

the form that meep wants for a (complex/dispersive) dielectric is: e + omega^2 de / ( omega^2 - w^2 - i w gamma).

the drude approximation goes as e - wp^2 / (w^2 + i w gamma). so, to get them to agree set omega to be some small number (much smaller than any w or sqrt(w gamma), so you can ignore it in the denominator and then you get e - omega^2 de / (w^2 + i w^2 gamma), so finally set wp^2 = omega^2 de, or de = wp^2 / omega^2.

finally, meep's units of w is always 2 pi c / a, where a is some characteristic length in your problem. so omega and gamma must also be in those units.
meep-discuss mailing list

Reply via email to