Bug report (NEB calculations)

[deepak_seth]

Code: Select all

Electronic structure
  PREC    = ACCURATE
  ADDGRID = .TRUE.
  LREAL   = Auto
  ENCUT   = 520
  ISYM    = 0
  ISPIN   = 2
  MAGMOM  = 126*0.0 36*0.0 17*0.0 1*0.0 1*0.0

Electronic minimization
  IALGO   = 48
  EDIFF   = 1e-5
  NELM    = 1000
  MAXMIX  = 80
  LMAXMIX = 4
  ISMEAR  = 0
  SIGMA = 0.05

Density functional
  GGA     = PE
  LDAU    = .TRUE.
  LDAUTYPE = 2
  LDAUL =  -1 -1 2 2 -1     ##order: O Nb Ti Ni Li
  LDAUU =  0.0 0.0 4.2 6.2 0.0
  LDAUJ =  0. 0. 0. 0. 0.
  LDAUPRINT = 0
  LASPH = .TRUE.

Job parameters
##ISTART  = 1
  LWAVE  = .FALSE.

NEB
   ICHAIN  = 0
   IMAGES  = 8
   ISPRING = -5
   LCLIMB  = .TRUE.
   LDNEB   = .TRUE.

Ionic relaxation
  ISIF    = 0
  EDIFFG  = -0.1
  NSW     = 1000
  IBRION  = 1
  POTIM   = 0.1

Parallelization
  NSIM    = 4
  NPAR    = 2
  KPAR    = 2
  LPLANE  = .TRUE.
  LSCALU  = .FALSE

Other parameters

I am trying NEB calculations of an oxide material with the above INCAR setting. I am getting a bug error as follows. This is the first time I am getting this error. Please help and resolve this issue.

Code: Select all

 -----------------------------------------------------------------------------
|                     _     ____    _    _    _____     _                     |
|                    | |   |  _ \  | |  | |  / ____|   | |                    |
|                    | |   | |_) | | |  | | | |  __    | |                    |
|                    |_|   |  _ <  | |  | | | | |_ |   |_|                    |
|                     _    | |_) | | |__| | | |__| |    _                     |
|                    (_)   |____/   \____/   \_____|   (_)                    |
|                                                                             |
|     internal error in: radial.F  at line: 757                               |
|                                                                             |
|     2 -0.543 -0.431 73.451 -47.228 43.514                                   |
|      2 -0.543 -0.431 73.452 -47.229 43.515                                  |
|      internal error in RAD_INT: RHOPS /= RHOAE                              |
|                                                                             |
|     If you are not a developer, you should not encounter this problem.      |
|     Please submit a bug report.                                             |
|                                                                             |
 -----------------------------------------------------------------------------

--------------------------------------------------------------------------
Sorry!  You were supposed to get help about:
    mpi-abort
But I couldn't open the help file:
    /proj/nv/libraries/Linux_x86_64/24.1/openmpi4-12/244921-rel-1/comm_libs/12.3/openmpi4/openmpi-4.1.5/share/openmpi/help-mpi-api.txt: No such file or directory.  Sorry!
--------------------------------------------------------------------------

[max_liebetreu]

Hello,

Welcome to the VASP Forum, and thank you for reaching out to us!

In order to investigate the bug you are seeing, please provide a minimal reproducible example, including all files necessary for performing your computation as well as the OUTCAR. Please attach your zipped computation folder as an attachment to your reply.

Best regards,

[deepak_seth]

Hello Max / VASP devs,

Thank you so much for the quick reply.

Please find the attached zipped computational folder for the NEB run.

####Attached (zip: nebrun_deepak.zip) — please find the following files:

INCAR (as given above)
KPOINTS
POTCAR
POSCAR files for images 01 to 08 (all NEB images in separate folders, consistent ordering) and for folder 00 and 09 (optimized POSCARs for the initial reactant and final product)
OUTCAR (complete OUTCAR from the failed NEB run)
pbsvaspsub_2_new.sh (PBS submission script, including command line and mpirun/launcher invocation used)
stderr_file: is the standard error file

####VASP Module and Version Detail:
$ module show apps/VASP/6.2.0.VTST/impi2020
-------------------------------------------------------------------
/home/soft/modules/apps/VASP/6.2.0.VTST/impi2020:

module-whatis VASP 6.2.0 (CPU, GPU)+ VTST with Intel 2020u4 CUDA 11.0 (CUDA C-Port)
module load suite/intel/parallelStudio/2020
module load compiler/cuda/11.0/compilervars
prepend-path PATH /home/soft/centOS/apps/licenseSoft/VASP/6.2.0.VTST.intel2020u4.cuda11.0/bin
prepend-path PATH /home/soft/centOS/apps/licenseSoft/VASP/6.2.0.VTST.intel2020u4.cuda11.0/vtstscripts-967
setenv OMP_NUM_THREADS 1
-------------------------------------------------------------------

I would appreciate it if you could resolve this issue.

[max_liebetreu]

Hello,

I am currently checking if I can reproduce your error without VTST.

In the meantime - are you aware that there's on pair of atoms in your setup (see 04/OUTCAR - nearest neighbor list) with an extremely low distance of 0.26? That might very well be the issue.

Best regards,

[max_liebetreu]

Hello,

The error appears without VTST as well. We do believe it is due to that one extremely close pair of atoms. Could you double-check the positions and get back to us on whether that was intentional or not, and if the error persists if that distance is corrected?

Best regards,

[deepak_seth]

Hello Max,

Thank you so much for checking this. You are right, that extremely close atomic distance (0.26 Å) was not intentional. I will correct the positions and re-run the NEB to see if the error persists. I’ll update you with the results.

Best regards,
Deepak

[deepak_seth]

Hello Max,

I checked the atomic distances and corrected my material model. After making the changes, the NEB calculation is now running smoothly without any errors.

Thank you very much for your support.

Best regards,
Deepak

[max_liebetreu]

Hello Deepak,

Thank you for letting us know! That's great to hear.
Glad it works now, and happy to help. I will be locking the topic as resolved soon, then. Have a good day!

Best regards,