As far as I know, there remains the problem of the mode shape
generally being different for different frequencies. In some specific
cases, such as hollow metallic waveguide, the mode cross-section
remains the same, so it is easy to excite/detect the first mode
selectively at broad frequency range.
The frequency dependence isn't a major problem for computing output power in
each mode, since generally you only need the power at a small number of
frequencies (a few dozen to a few hundred), and Meep can compute the explicit
Fourier transforms of the fields in the flux plane at these
2 matches
Mail list logo
