From: Thomas Brumme <thomas.brumme@na...> - 2012-02-07 12:43:30
first of all, thank you for this awesome piece of code!
I'm actually trying to use WavePacket to simulated the wavepacket
propagation through a 3d potential I get from a plain DFT calculation.
After a few problems with the interpolation (I essentially had to change
the interpolation method from "interp3" to "griddata3", using the
nearest neighbor method) it now seems to work. However, there is one
problem I encountered during visualizing the result in the "reduced"
plot form. I start my calculations with a momentum of [0 0 2] (particle
moving in positive z-direction), but the plots always show gaussian
packets with the center near a momentum of [0 0 -2]. But this can only
be a problem of the visualization since the code writes correctly [0 0
2] in the console...
Furthermore, another strange thing I had... I used periodic boundary
conditions and a test potential. The cell has 50x50x50 points. Depending
on the conversion factor I can make any unit cell size out of this
(cubic of course). And I changed the conversion factor. However, for
each different value of unit cell size, the wavepacket spreads very
rapidly and after a few time steps the center of the packet is in the
center of the cell and it only gets broadened. I should use a correlated
wavepacket in this case to get more reasonable results, or?
One last question: Is there a easy way to print out the gradient of the
potential at the center of the wavepacket to get the electric field at
this point? Can you give me maybe at least a hint in which part of the
code I would have to add the corresponding matlab expression? (i.e. in
which variable is the center of the wavepacket stored and where is it
Thank you in advance
Dipl.-Nat. Thomas Brumme
Institute for Materials Science and Max Bergmann Center of Biomaterials
Dresden University of Technology
01069 Dresden, Germany
Tel: +49 (0)351 463 31408