ERROR: SBESSELITER : nicht konvergent

Message

baknoo · #1 Post by **baknoo** » Thu Jun 28, 2007 5:45 am

Dear VASP expert

My VASP job stop with the message
"ERROR: SBESSELITER : nicht konvergent".
What does it mean ? and How to resolve this problem ?
Below I attach the standard output written in "vasp.out".

Thank you very much in advance for your helpful answer.

running on 16 nodes
distr: one band on 16 nodes, 1 groups
vasp.4.6.26 15Dec04 complex
POSCAR found : 4 types and 340 ions
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent

#2 Post by **admin** » Mon Jul 02, 2007 12:06 pm

the error message says that the spherical bessel function cannot be calculated to the required accuracy. Please give some more details about your calculation.

ypipon · #3 Post by **ypipon** » Fri Dec 05, 2008 2:14 pm

Dear VASP Expert,

I had the same problem a few moments ago. I am currently studying a TiN matrix. VASP is ok with 2, 8, 16, 32, 64 atoms but not with 128 atoms.
Have you a hint for me ?

Thank you by advance.

#4 Post by **admin** » Fri Dec 05, 2008 2:37 pm

sorry, the info you give is rather poor. Have you rescaled your unit cells properly?

fabrizio.cossu · #5 Post by **fabrizio.cossu** » Mon Feb 04, 2013 2:17 pm

I get the same error message.
my input has:

# initialization of the run
ISTART = 1
ICHARG = 1
ISIF = 2

# dimension of arrays
# NGX = 27
# NGY = 27
# NGZ = 115

NBANDS = 500

# magnetic parameters
ISPIN = 2
# MAGMOM = 16*0.0 16*0.0 60*0.0 4*4.0 4*0.0

# electronic loop
LREAL = On
ENCUT = 500
IALGO = 38
PREC = Medium
# NSIM =
EDIFF = 1E-4
NELMIN = 7

# ionic loop
IBRION = -1
# SMASS = 1.0
# NSW = 10
# EDIFFG = -1E-1
# POTIM = 0.2

# storage of info
# RWIGS = 4.040 2.650 2.500 1.550
LWAVE = .TRUE.
LCHARG = .TRUE.

# DoS calculation
LORBIT = 11
ISMEAR = -5
SIGMA = 0.01
NEDOS = 1001
EMIN = -5
EMAX = 15

# +U module
LDAU = .TRUE.
LDAUTYPE = 1
LDAUL = -1 2 -1 2 -1
LDAUU = 0.00 5.00 0.00 4.00 0.00
LDAUJ = 0.00 0.64 0.00 0.98 0.00
LDAUPRINT = 2

---

with KPOINTS

Automatic mesh
0
Gamma
4 4 2
0. 0. 0.
---
and i have 100 atoms.
I'm using the WAVECAR and CHGCAR from a relaxation run, which has been completed successfully.

fabrizio.cossu · #6 Post by **fabrizio.cossu** » Mon Feb 04, 2013 3:16 pm

I also did few tests:
1) no restart, with same KPOINTS: 4x4x2
2) restart, as for the one giving error, with less KPOINTS: 4x4x1.

In both cases i don't get the message

plasienka · #7 Post by **plasienka** » Fri Oct 11, 2013 10:50 am

Dear VASP expert

My VASP job stop with the message
"ERROR: SBESSELITER : nicht konvergent".
What does it mean ? and How to resolve this problem ?
Below I attach the standard output written in "vasp.out".

Thank you very much in advance for your helpful answer.

running on 16 nodes
distr: one band on 16 nodes, 1 groups
vasp.4.6.26 15Dec04 complex
POSCAR found : 4 types and 340 ions
ERROR: SBESSELITER : nicht konvergent
ERROR: SBESSELITER : nicht konvergent

According to my experience, to get rid off this error, I had to decrease the cutoff.

SKM · #8 Post by **SKM** » Fri Aug 21, 2020 9:12 am

hi

@plasienka;

According to my experience, to get rid off this error, I had to decrease the cutoff.

but in my case i am testing convergence for optimum ENCUT.
it stated at higher values of 650eV and 675eV for a simple primitive cell of perovskite.

As per your answer, "i cannot go beyond 650eV" in my case?
what is the solution actually. ? this thread stopped in between.
kindly anyone will answer this?
Regards

#9 Post by **henrique_miranda** » Wed Sep 30, 2020 1:06 pm

Note that the answer suggesting to decrease the cutoff comes from a user reporting his personal experience.
This does not mean necessarily that the correct way to solve this issue is by reducing the cutoff.

Also as you can see above, this thread finally did not get an answer because the information provided by the users is very incomplete.
If you have the same issue please post INCAR, POSCAR, KPOINTS, the first line of the potcar files (not the POTCAR files themselves) as well as the OUTCAR files.
The calculation should be easy to run (i.e. not too many atoms, not too many k-points and so on) so that the issue is easy to reproduce in a short run.

SKM · #10 Post by **SKM** » Sat Nov 21, 2020 11:27 am

hi Admin,

I have just got similar issue. This is the same system which i posted an issue in the topic "SCF Convergence for a big system" forum/viewtopic.php?f=4&t=10889.
I am waiting for your reply on the above topic. Meanwhile, i started ENCUT convergence on the same system and the following issue occured.

For the same POSCAR, POTCAR, KPOINTS, and for only difference in ENCUT value in INCAR, the system thowing error for ENCUT 650eV, 700eV, but run smoothly for ENCUT 600eV.
attached all three cases tar files.
kindly see and suggest what is the issue and how to resolve.
Regards

#11 Post by **henrique_miranda** » Tue Nov 24, 2020 8:51 am

I looked at your output files and I've only encountered this error:

Code: Select all

 REAL_OPTLAY: internal error (1)     7099239     7099642

This is different from the error being discussed in this thread.

You might find the answer to your issue here:
forum/viewtopic.php?f=4&t=5354

My Community