Problems restarting with the CHGCAR-file only

[boris]

Hello

I am encoutering problems while trying to restart my jobs using only the CHGCAR-file.

Below is what is written in the VASP manual :
[quote="VASP Manual"]
The MAGMOM tag is used only, if the CHGCAR-file holds no information on the magnetisation density, and if the initial charge density is not calculated from the wavefunctions supplied in the WAVECAR-file
[/quote]

What I understand from this is that MAGMOM tag is useful with :
1. ISPIN = 2 ; ISTART = 0 ; ICHARG = 2. Job starting from scratch.
2. ISPIN = 2 ; ISTART = 0 ; ICHARG = 1. But only if the CHGCAR-file has been created with a ISPIN = 1 job.

I understand as well that when restarting from a magnetic CHGCAR-file (created with a ISPIN = 2 job), I have to comment the MAGMOM line.

Here is my problem.

I made a calculation using :
ISPIN = 2
ISTART = 0
ICHARG = 2
MAGMOM = 2 -2 0 0 0 0

The calculation is running well, and I want to use the CHGCAR-file obtained in this calculation to make a static calculation in order to have an accurate DOS-file.

So I'm using :
ISPIN = 2
ISTART = 0
ICHARG = 1
#MAGMOM = 2 -2 0 0 0 0

With these settings, VASP converges towards a non magnetic configuration (all magnetic moments are zero, on each component), although it writes in the OUTCAR-file that it has successfully read the charge density from the CHGCAR-file.

Did I miss something ?

I'm using PAW and GGA+U.

Thanks for your answers.

[admin]

if you just want to do a (post-processing) DOS run, please set ICHARG=11 (otherwise, vasp will start another scf procedure, possibly relaxing to a different -low spin- solution (especially, if no initial moments are set).
Please try ICHARG=11