My Community

Dear All:
I am having problems computing the phonon frequencies in VASP version 5.2 with IBRION=6 The structure is converged with IBRION=2 but a run to obtain the phonons does not stop after energy convergence on wavefunctions and calculate the phonon frequencies. Any pointers would be useful. The INCAR and POSCAR files are below.
Trevor


*****************
INCAR

System=a-b-c, LDA+U, symm turned off, FM magnetic order
PREC= Accurate
ADDGRID = .TRUE.

Parallel Stuff
NPAR=8

ENCUT= 500.00 eV
ISPIN= 2 !spin polarized calculation
LORBIT=11
MAGMOM= 4 4 4 4 16*0 !make moment larger expected val.


ISTART=1
ICHARG=0
IBRION= 6 ! 2=conj grad and 1=quasi-Newton (variable metric) algorithm
NFREE=4
POTIM=0.015 !displacement of atoms in Ang.
LCHARG= .FALSE.
LWAVE= .FALSE.
LREAL=.FALSE. !needed for good convergence
ISYM=0 !switch Off symmetry (PAW potential)
IALGO= 38
EDIFF =1.0E-08


Ionic Relaxation (coarse for structural opt only)
NSW= 1
ISIF= 3 ! adjust abc xyz
EDIFFG= -5.0E-04 !5.0E-04 eV/A





************************
POSCAR
LDA+U, a, b, c, sy
1.000000000000000
5.2827497172433660 0.0000000000000000 0.0000000000000000
0.0000000000000000 5.8096041972536630 0.0000000000000000
0.0000000000000000 0.0000000000000000 7.5181314824947640
4 12 4
SELECTIVE DYNAMICS
Direct
0.5000000000000000 0.0000000000000000 0.0000000000000000 T T T
0.0000000000000000 0.5000000000000000 0.0000000000000000 T T T
0.0000000000000000 0.5000000000000000 0.5000000000000000 T T T
0.5000000000000000 0.0000000000000000 0.5000000000000000 T T T
0.1289637528571407 0.4505034259741735 0.2500000000000000 T T T
0.3710362471428598 0.9505034259741736 0.2500000000000000 T T T
0.6289637528571407 0.0494965740258261 0.7500000000000000 T T T
0.8710362471428594 0.5494965740258264 0.7500000000000000 T T T
0.6968524002712170 0.3255352627755137 0.0615065384729744 T T T
0.8031475997287830 0.8255352627755212 0.0615065384729744 T T T
0.1968524002712090 0.1744647372244789 0.5615065384729743 T T T
0.3031475997287906 0.6744647372244787 0.5615065384729743 T T T
0.3031475997287906 0.6744647372244787 0.9384934615270257 T T T
0.1968524002712090 0.1744647372244789 0.9384934615270257 T T T
0.8031475997287830 0.8255352627755212 0.4384934615270261 T T T
0.6968524002712170 0.3255352627755137 0.4384934615270261 T T T
0.9787805816886825 0.0823796520245213 0.2500000000000000 T T T
0.5212194183113175 0.5823796520245219 0.2500000000000000 T T T
0.4787805816886818 0.4176203479754792 0.7500000000000000 T T T
0.0212194183113179 0.9176203479754781 0.7500000000000000 T T T

please note that the number of displacements which have to be calculated if all atoms are allowed to vibrate and NFREE = 4 will be probably be extremely large, even is symmetry is taken into account (IBRION=6) because as far as I see at first glance from your POSCAR, the symmetry is probably very low. There is an electronic scf calculation done for each of the displacement to set up the Hessian matrix, only displacements which are equivalent due to one of the symmerty operations of the lattice are not calculated explicitely