6
$\begingroup$

I'm encountering convergence issues in my VASP relaxation calculations, specifically related to the message 'BRIONS problems: POTIM should be increased.' I've tried adjusting POTIM, but I'm still facing difficulties achieving convergence. Are there any other parameters or strategies I should consider to improve convergence in my VASP simulations?

ISTART = 0 ; ICHARG=2
ENCUT  =   500
ISMEAR =    1  ; SIGMA = 0.005
ISPIN = 2
ISIF =3
EDIFF = 1E-8
EDIFFG = -0.0001
NSW = 200
IBRION = 1
POTIM = 3
ISYM = 1
$\endgroup$
1
  • 1
    $\begingroup$ the course of action here may depend on VASP version. What is your VASP version, and what other settings with respect to IBRION and POTIM have you tried? At a first glance, your POTIM looks huge for the extremely tight convergence you seek, I would use 0.02 or something like that. See also the note on how more recent VASP versions handle POTIM being too large: vasp.at/wiki/index.php/POTIM $\endgroup$ Commented May 6 at 17:12

1 Answer 1

5
$\begingroup$

The problem comes up because you are using such a small EDIFFG, first consider if you really need this small of an EDIFFG. If so, you can try the following.

  • Lower POTIM, likely <0.01
  • Use an adaptive step size method such as FIRE in VTST
  • Tighten your electronic convergence, the grid may play a role here. ADDGRID=True may help and/or PREC=Accurate
  • If the cell shape is good, you can try ISIF=8 as seen here
  • Remove symmetry, at this small of an EDIFFG it is possible the symmetry won't allow a smaller force. ISYM=0
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .