I would like to know the cause of the failure in my calculation.
I am trying to get the data of an structure of liquid Al2O3, so I have recently calculated the MD simulation by using VASP. However, a few days after I threw the job I found there appeared an error message in final line of the OUTCAR such as
"Error EDDDAV: Call to ZHEGV failed. Returncode = 11 1 12"
which indicates there is something wrong with the matrix diagonalization.
The error showed up in "Iteration 13 (1)", so I guess the error doesn't derive from the initial structure of the Al2O3. The INCAR I have applied is
System =Al2O3
GGA = PE
ISPIN = 2
MAGMOM=160*0 1
RWIGS= 0.820 1.402 0.370
ISMEAR = 0
SIGMA=0.05
ENCUT=500
LWAVE=.FALSE.
LCHARG=.FALSE.
IBRION=0
MDALGO=2
ISIF=2
SMASS=1.0
TEBEG=4000
NSW=20000
POTIM=5.0
of which the parameters of the temperature(TEBEG) and the time length(NSW*POTIM) are what I quoted from a previous research(J. Chem. Phys. 153, 164119 (2020)).
I am a graduate student who don't understand the DFT or VASP very well. Addtionally I am a student in Japan, so I would like you to pardon my terrible English.
Hi Kodai Terashita
This problem could occur due to multiple reasons. Please try the following options to solve it.
- Check if your input geometry is reasonable. Look at the POSCAR and check if any atoms are overlapping. Overlapping of atoms or them being very close to each other in the POSCAR is seen as the cause of this error.
- Use a different algorithm for diagonalization. Use ALGO tag to set it.
- Try different number of cores (Ref: https://www.vasp.at/forum/viewtopic.php?f=2&t=10409&p=10434)
Best
- Badges
- Science topic