Page 1 of 1

ISMEAR

Posted: Thu Oct 13, 2011 8:34 am
by vasp_user
I am trying to get the "VASP theory value" for the lattice constant for hcp Mg. My plan is to do it with "production level" convergence. So first I found that a 16x16x16 Monkhorst-Pack grid is converged and now I am going to do the energy cut off convergence before I start searching for the optimal lattice parameter.

I got to thinking... since I am only doing energy cut off convergence now and not moving the ions at all I can use ISMEAR=-5. But later I might want to do be able to have ISMEAR = 1 when I am searching for the optimal lattice parameter. Will my energy cut off that I found with ISMEAR=-5 still be valid when I move on to searching for the optimal lattice parameter and set ISMEAR = 1?

The answer would be nice... but a few sentences on the thought process that leads to the answer would be really really great.

Regards,
vasp_user

ISMEAR

Posted: Wed Oct 19, 2011 1:55 am
by vasp_user
Looks like ISMAR=-5 should be used with a gamma centered k-mesh.

ISMEAR

Posted: Tue Oct 25, 2011 4:46 am
by Dr Vitaly V. G. Chaban
[quote author=-5 still be valid
[/quote]

Yes, it should be.