On Jan 4, 2012, at 6:08 PM, Jun Liu wrote: > Dear Dr. Baroni and other forum members, > > Thanks for your kind and patient response on my question. I am glad to know > that there is a way to cheat QE to do a free electron gas calculation by > giving a null pseudo-potential. I guess this "null" means replacing every > number in the pp file with zero. If the case, then I might ask a bit further > on my question.
well, you have to understand what the various "zeroes" mean, and be sure that what you do makes sense. for this, I am afraid, you will need to dig a bit into the UPF format. not sure the effort would be worth the result. I would rather suggest to try first what Lorenzo Paulatto suggested, and then, if anything else fails, it may be simpler to modify the code than to fudge the .upf files ... > With a free electron gas system, plane wave expansion would be enough. But > what if I want to replace the plane wave basis with the plane wave + some > local orbital basis and do the calculation. not sure I understand why with the zillions of interesting physics problems to solve that exist out there you want to solve a trivial problem with a very complicated method ... in any case, mixed basis sets are not implemented in any of the QE codes ... > Although a weird request, this should be able to be done in principle. But > the issue is how to introduce the local orbits into the problem. Can I do > something like this? yes, but not with QE > Choose some pp giving the local orbitals to be included in the calculation, > and set in the input pp file the local/nonlocal potential and the pseudo/AE > core charge to vanish, to mimic the fact that there is no ion existing in the > system. Then what is left in the pp file might render QE to do some type of > plane wave+local orbital calculation. Please let me know whether this idea > works and whether these modifications are sufficient for the purpose. I do not understand what the idea is. I am afraid you are confusing the localized nature of *some* basis sets (including, e.g., gaussians, or atomic orbitals) with the atomic wavefunctions that come with the .upf file and that are *NOT* used as basis functions. be sure you understand what US PP, PAW, and the like work ... > I might here ask a very fundamental question. The reason we choose the > sophisticated PAW treatment by introducing projector/pseudo/AE wavefunction > directly relies on the non-smoothness nature of the valence orbital itself, > but not on a specific ion to be treated. That is to say, we don't need to > know what is behind the scene when deciding whether PAW treatment is needed > given a specific valence orbital in the pp file. Am I right? you always need to know "what is behind the scene": that's exactly the purpose of Science ... Stefano B --- Stefano Baroni - SISSA & DEMOCRITOS National Simulation Center - Trieste http://stefano.baroni.me [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype) La morale est une logique de l'action comme la logique est une morale de la pens?e - Jean Piaget -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20120104/0e53cfc8/attachment-0001.htm
