Total local potential

[apple]

hi,
I have a question about how to calculate the total local potential. According to the vasp manual that this is done by placing a test charge with the norm 1 at each ion and calculating the volume integral (http://cms.mpi.univie.ac.at/vasp/vasp/node144.html). Does it mean that one must specify RWIGS in the INCAR file? What is that volume in the integral?
Also, does one have to run a static selfconsistent run to determine charge density first and then use this density (ICHARGE=1) in the non-consistent run to get the potential?
Any help is greatly appreciated.

[forsdan]

1) The manual clearly states that it is the average electrostatic potential at each atom that is evaluated by placing a test charge with the norm 1 at each ion and calculating the volume integral. Please do not confuse this with the total local potential.

2) The integration volume for evaluation of the average electrostatic potential is decided by the spatial distribution of the test charge which in turn is determined by ENAUG. So RWIGS has no impact in this context. Please also note that test charge radius is written in the OUTCAR file where the average electrostatic potential at each ion are given.

3) The total local potential in the computation cell is written to the LOCPOT file (if LVTOT = .TRUE.) at the end of the self-consistency cycle. By default only the electrostatic part is written but it is also possible to add the exchange-correlation potential (see the manual for the LOCPOT file, although note that for VASP5.2 the line to comment out is different).

Cheers,
/Dan




<span class='smallblacktext'>[ Edited Wed May 12 2010, 10:53PM ]</span>

[apple]

Dan: Thanks a lot for your explaination. It really helped!