Hello
I don't know why when I do the relaxation of my orthorombic structure "CH3NH3PbI3" there is not a change in the lattice parameters for example:
before the optimization:
- a=8.3541612625
- b=12.6614093781
- c=9.0433378220
after optimisation
- a=8.3541612624999999
- b=12.6614093781000001
- c=9.0433378219999998
I know what the problem is ????????????
here are the input and output files poscar , incar , k-point and contcar ,
Increase the NSW value (say 1000) and check EDIFFG (https://www.vasp.at/wiki/index.php/EDIFFG).
Can you share the OUTCAR file too for a more detailed view of the problem? Soufiane Elhadfi
As was pointed out, the current EDIFFG is extremely strict convergence criterion. The way is written, the convergence criterion is for the forces (the minus sign), and usually -0.01 or -0.001 are used. Hence -1E-8 is extremely high precision, almost impossible to reach for your system. Start with -0.01, and if you need stricter criterion increase it to -0.001.
Also, NELM of 300 seems like a lot of SCF steps, I would not use more than 100.
Lastly, you can remove the symmetrization of the charge density by putting ISYM=0 and see if that solves your problem.
- Badges
- Science topic
- Similar topics
- Filing