Doubts about total drift

[alberto.trj]

Hi people, I have some doubts about the total drift:

1) How is the total drift calculated?
2) What are the effects of a big drfft on the simulations?
3) Are calculations (say band structure, for example) with a total drift larger than the forces on the ions reliable?

Thanks in advance

[admin]

1) the drift is calculated from the sum of the forces on the center of mass
2) this depends:
there may be 2 reasons for a large total drift:
a) if you have a slab or a cluster and you do not keep at least one atom fixed, the whole slab or molecule may start drifting through the vacuum (--> please use the selective dynamics option in POSCAR in this case and keep at least one atom or layer at a fixed position)
b) it may be due to 'aliasing' - errors, caused by too small basis sets (please have a look at accuracy.pdf at
http://cms.mpi.univie.ac.at/vasp-workshop
for a detailled explanation
3) it depends: if only 2a) is the reason (and the basis stets are sufficiently large), the relative forces between the atoms are correct, if the basis stets are too small, the forces between the may be not accurate.