Error in Dynamical Matrix run

[sinfire]

Hi all

I'm trying run a dynamical matrix run to get the ZPE corrections.

When running the job, I'm getting errors with NaNs appearing in the output.

I'm running the job with 9 DOFs, 3 images with NSW=4.

Could anyone explain why this error appears and suggest a way to avoid this error?

I check the job does fine for other systems and parameter settings,
so I assume this is an error that can be fixed by tuning the calculation system or setting

I'll attach the files I used for the calculation.

Thanks

M.-C.


Error message in OUTCAR:
===================================
....
DYMAT: LEAVING
stress matrix after NEB project (eV)
-7.62932 -0.25918 0.13745
-0.25918 -6.93207 0.40593
0.13745 0.40593 -1.63994
FORCES: max atom, RMS 0.395335 0.120593
FORCE total and by dimension 0.525654 0.324304
Stress total and by dimension 10.461900 7.629325
Steepest descent step on ions:
trial-energy change: -0.000644 1 .order 0.000000 0.000000 0.000000
(g-gl).g = 0.000E+00 g.g = 0.000E+00 gl.gl = 0.000E+00
g(Force) = 0.000E+00 g(Stress)= 0.000E+00 ortho = 0.000E+00
gamma = 0.00000
trial = 1.00000
opt step = NaN (harmonic = NaN) maximal distance = NaN
next E = NaN (d E = NaN)


--------------------------------------------------------------------------------------------------------


WAVPRE: cpu time 1.3037: real time 1.0094
FEWALD: cpu time 0.0009: real time 0.0009

[sinfire]

OK I think I solved the problem.

It turned out that I didn't read the dynmat page more carefully.


I had to set IBRION=3 but my calculation was set as IBRION=2.

Strange that other jobs worked fine though.