VASP 6.4.3 Initialization

Queries about input and output files, running specific calculations, etc.


Moderators: Global Moderator, Moderator

Post Reply
Message
Author
hashan_peiris
Newbie
Newbie
Posts: 1
Joined: Tue Jan 05, 2021 8:54 pm

VASP 6.4.3 Initialization

#1 Post by hashan_peiris » Thu Oct 03, 2024 10:23 am

I started running test calculations using VASP 6.4.3 and I noticed that the line "LDA part:" in the OUTCAR during the initialization changed between VASP 6.4.2 and 6.4.3 from "LDA part: xc-table for Pade appr. of Perdew" to "LDA part: (Slater+PW92)."

I noticed the addition of the XC tag in the new version, but I could not make it go back to the old scheme. Is this intended behaviour?

Here is my INCAR:

#GGA = PE
#XC = GGA

ISMEAR = 0
SIGMA = 0.20

IVDW=12
LASPH = TRUE
ENCUT = 450
PREC = Normal

NELM = 500
NELMIN = 5
EDIFF = 1E-5

IBRION = 2
ISIF = 2
NSW = 10000
EDIFFG = -0.01

ISYM = 0
ALGO = Fast
LREAL = Auto

LCHARG = False
NCORE = 10

The INCAR and the POTCAR remained the same. I would like to keep this the same for continuity in my calculations.

From my understanding, these two approaches aren't the same.

Thanks!


martin.schlipf
Global Moderator
Global Moderator
Posts: 542
Joined: Fri Nov 08, 2019 7:18 am

Re: VASP 6.4.3 Initialization

#2 Post by martin.schlipf » Thu Oct 03, 2024 12:16 pm

There has been a change to the treatment of xc potential in the code to allow for a better interface with libxc and more consistent treatment of the xc terms. It seems that in this refactoring the label of this output was changed. As far as I could see from the code, this should not affect the results though. Do you see changes to the output when you run the two different version on your system?

If there are some small changes and you want to know whether they originate from this part of the code, you can look into the routine setex.F. At the end of the routine SET_EX_TABLE there is a statement IF (.FALSE.) THEN. If you replace this line with IF (.TRUE.) THEN you can write the XC table to file and compare it directly.

Please let us know, if you spot any differences.

Martin Schlipf
VASP developer


Post Reply