Page 1 of 1

Convergence criteria for stress?

Posted: Thu Dec 18, 2014 2:20 am
by xamnidar
In cases where EDIFFG < 0, how does VASP determine the convergence criteria for the relaxation of the lattice parameters? Is the value of EDIFFG converted to a stress?

Re: Convergence criteria for stress?

Posted: Wed Sep 02, 2015 11:32 pm
by xamnidar
EDIFFG gives the stress convergence criteria in units of eV/(volume per atom), if my reading of the code is correct. Volume per atom here means the cell volume divided by the number of atoms in the cell.

Note that in the OUTCAR, the FORCE on cell section gives the total stress in units of eV/(cell volume). Dividing these values by the number of atoms will yield the stress in units of eV/(volume per atom).