My Community

Can we change the mass of some atom to be smaller or bigger than from POTCAR, because when I read from VASP manual we can change the mass by set the POMASS, but when I try to change it to be 10x smaller and 10x bigger theres no different in total energy at all (I also change the value from the POTCAR, so vasp will calculate it). I'm not too sure, but mass should be change the kinetic energy and would also produce a different total energy

Can we really change the mass? or not?

If you are doing a relaxation then I would imagine the following is happening:
1) You calculate the electronic energy at the initial configuration of the ions (atoms).
2) VASP calculates a force on the atoms after that initial electronic calculation converges.
3) VASP usus the forces to update the atomic positions
4) VASP does another electronic calculation etc.

Now the above just keeps repeating until the forces on the atoms are below the tolerance you set. For my money I would say that -each time through the above process- the forces just will not move the atoms as far when you increase the mass (and the forces move the atoms further each time if you decrease the mass). But the run goes until the final configuration is within the tolerance you set... and so the energy -once everything is within that tolerance you set- is what it is... it will not change just because of the size of the steps that were taken on the way there. Does that answer your question?

I already think about this one, and I remember that its related to Born-Oppenheimer approximation that they ignored the nuclear movement and that would make the kinetik energy for nuclear disappear, but the things that makes me feel strange is, "Why theres POMASS parameter if it will not change anything?", ^_^

[quote="b_lack"]
I already think about this one, and I remember that its related to Born-Oppenheimer approximation that they ignored the nuclear movement and that would make the kinetik energy for nuclear disappear, but the things that makes me feel strange is, "Why theres POMASS parameter if it will not change anything?", ^_^
[/quote]
The Born-Oppenheimer approximation decouples the equation of motion for the electrons and the ions. It does not mean that the nuclear movement is ignored. The pomass is needed for molecular dynamics simulations and phonon claculations.

Oke, now it start to make sense, because I never done any MD and Phonon so I didnt know about this one

Thanks tlchan