LOCPOT, nonlocal part
Posted: Sun May 03, 2015 11:55 pm
Hi, I have a question regarding the total local potential: Does LOCPOT contain the total (Hartree+ionic+xc) sc local pseudo potential or is it an actual potential? My understanding is that in order to obtain a physically meaningful potential nonlocal contributions would need to be added (which cannot be simply added due to their operator nature). Is that correct? Is there a way to estimate the strength of nonlocal contributions from VASP output? Or are these nonlocal contributions typically very small compared to the local part?