On Sun, 15 Apr 2007, Lydie Ferrier wrote: > I've got problems when computing the complex frequency of modes at > Gamma-point : when the modes have high Q-factor (small imaginary part), > there is no problem and i get the bands. But when the modes at > gamma-point have very low Q-factor (and thus large imaginary part) MEEP > do not compute the complex frequency around Gamma-point. The band begin > at a k value quite "far" from gamma when the imaginary part of the > frequency becomes smaller... Is there any "trick" to avoid the problem??
First, you should realize that the question you are asking may not be very well defined. Once Q drops below 30 or so, the whole notion of a leaky mode with a "complex frequency", which is only a near-field approximation in any case, becomes problematic. Second, as a practical matter, harminv has a difficult time getting accurate results for very low Q modes, because they decay away too quickly after the source turns off. For these reasons, when Meep calls harminv it is pre-set to discard any modes with Q < 50. You can change this by editing the Q_thresh default in the do_harminv prototype in src/meep.hpp, but for the above reasons you still may not get satisfactory results. Steven _______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
