Hi everyone!
I am doing phonon calculations of a 2X2X2 supercell of a 16 atom unit cell using DFPT (IBRION=8) in VASP. So, there are 64 atoms in total and it is a tetragonal crystal structure. Is it compulsory to invoke ISYM tag for doing phonon calculations in VASP? When I included ISYM=0 (idea obtained from the tutorials), it started calculating 192 degrees of freedom (64X3=192). However, when I did not mention about ISYM tag, only 8 degrees of freedom were calculated. I do not understand which one is correct.
Secondly, if I go ahead with the calculation of 192 degrees of freedom, the job does not finish even using GPU in a single go, due to its limited wall time. Is there any way to resume the phonon calculations in VASP or doing phonons with finite differences method is also the same?
Lastly, I would like to know if the smearing methods (ISMEAR) and its sigma value and the k-points could help in fast convergence in case of phonon calculations?
Could somebody please enlighten me on this so as to proceed further.
Thanks in advance!