Hello MPB specialist,
I've been working for sometime on reading the MPB fields (from hdf5 files), compared with the eigenvectors decomposition. I stumbled upon a problem that caused me to raised the question in the title.
I refer first to the 2D problem, say TE polarization where the magnetic field H has a single component in the direction perpendicular to the PhC plane - this will be the z direction. the PhC lattice axes form a set of coordinates x' and y', which can be different than x and y. I realized that mpb performs its computations relative to x' and y' for the space points, and in kx' and ky' (which can also be different than x and y) for the k vectors and G vectors (the spatial harmonics in the periodic decomposition of the dielectric and fields functions).
Up till now, I thought that the coordinates alone are affected by the transition to the lattice axes, which can be treated later with mpb-data to return to real x,y,z space. However, I looked through the mpb source code and found that, given a magnetic field vector H, the electric displacement D field is computed using the function maxwell_compute_d_from_H which (according to the comments) compute (k+G)xH. So my question is: are the components of D given in kx' and ky' coordinates? or is it x' and y'?
Or perhaps if I take the fields hd5 file after I pass them to mpb-data, then E will be in kx' and ky' and E-new will be in x and y?
Following this I need to ask what goes on with 3D PhC lattice, which field components are given in reciprocal axes / lattice axes?
Thank you very much in advance,
Amnon Willinger
_______________________________________________ mpb-discuss mailing list mpb-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/mpb-discuss
