[Q-e-developers] Seek for Technological Help

[Jiqiang Li]

Dear developers,
     I am a user of Quantum Espresso and want to revise the code to add some new function on it. However, I have some puzzles with me.
I want to evaluate the elements of electron-phonon coupling matrix due to local potential. i.e. <f|dV_loc/dR|i>.
|i> is the initial state and |f> is the final one. I evaluate \sum {psic_f* psic_i dV_loc/dR } with finite differential method. The result is strange, so I want to clarify something detail:
1. I employed the Vrs in the SUBROUTINE electrons_scf(no_printou) after  CALL interpolate_vrs, is it the correct V_local I should take?
2. I followed the algorithm in SUBROUTINE sum_band to evaluate (psic_f* psic_i), Can I directly mutiply (psic_f* psic_i) by Vloc when I had done the tranformations on the several grids?
3. How long is recommended for the finite differential displacement for one atom? I find the displacements about 0.01 Angstrom behaved no linear relations for  dV_loc/dR, but 0.001 Angstrom did. However, I am afraid a too small displacement would be drown by the numerical noise.
4. Is there any advice on the calculation?
   Thank you for your consideration!
Kind regards!


Ji-Qiang Li
Department of Physics, Fudan University




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20150128/197af9de/attachment.html>