Error while calculating elastic tensor with IBRION 6, ISIF 3

[barshab]

Dear VASP Community,

I am trying to calculate the elastic tensor for Cr2N using ISIF = 3 and IBRION = 6 but I always stumble upon the following error:

Code: Select all

 -----------------------------------------------------------------------------
|                     _     ____    _    _    _____     _                     |
|                    | |   |  _ \  | |  | |  / ____|   | |                    |
|                    | |   | |_) | | |  | | | |  __    | |                    |
|                    |_|   |  _ <  | |  | | | | |_ |   |_|                    |
|                     _    | |_) | | |__| | | |__| |    _                     |
|                    (_)   |____/   \____/   \_____|   (_)                    |
|                                                                             |
|     internal error in: mkpoints_change.F  at line: 185                      |
|                                                                             |
|     internal ERROR in RE_READ_KPOINTS_RD: the new k-point set for the       |
|     reduced symmetry case                                                   |
|      does not contain all original k-points. Try to switch off symmetry     |
|                                                                             |
|     If you are not a developer, you should not encounter this problem.      |
|     Please submit a bug report.                                             |
|                                                                             |
 -----------------------------------------------------------------------------

I have tried using ISYM = 0 (with NCORE since it works with ISYM) and the calculation does starting running but the number of steps for the finite differences is around 444 and they never complete in the maximum time limit I have from the supercomputer.
The crystal structure is tetragonal. I am using a system of 72 atoms and I use 192 cores with KPAR = 2 for 120 hours (5 days) Using NCORE also returns an error so I cannot use it for this type of calculations. I use DFT+U with U = 4, J = 1 with ISMEAR = 1.

Will be grateful for any help or insights provided!
Thank you for taking the time to read this.

Best wishes,
Barsha.

MS: Edited the error message for readability.

[martin.schlipf]

Can you please provide your input files so that we can reproduce this issue?

[barshab]

Dear Martin,

Thank you so much for getting back to me.

The INCAR is as follows:
!Initialization
System = Cr2N trigonal !Title
ISTART = 0 !New Job
ICHARG = 2 !how VASP constructs the initial charge density

!Parallelization
KPAR = 2
NCORE = 16
!NPAR = 4
LSCALU = .FALSE.
LPLANE = .TRUE.

!Electronic Optimization
ALGO = Normal
PREC = Accurate
EDIFF = 1E-06 !Force stop criteria
ENCUT = 770 !Cutoff Energy
LORBIT = 11 !Different Orbitals
LREAL = .FALSE. !Real Space Projection
LWAVE = .FALSE. !Wavecar is not written
NELM = 100 !The maximum number of electronic SC (self-consistency) steps


!Ionic Relaxation
EDIFFG = -1E-02 !stopping-criterion for IOM
IBRION = 2 !how the ions are updated and moved
NSW = 99 !number of ionic steps
ISIF = 3 !stress and relaxation
ISMEAR = 1 !Smearing, Methfessel-Paxton
SIGMA = 0.1 !Width of Smearing

!Magnetism
ISPIN = 2 !Spin polarized
MAGMOM = 48*5.0 24*0.6 !Magnetic Moment

!GGA+U
LMAXMIX = 4 !l-quantum number
LDAU = .TRUE.
LDAUTYPE = 2
LDAUL = 2 -1
LDAUU = 4.00 0.00
LDAUJ = 1.00 0.00
LDAUPRINT = 2

!Mixing
LASPH = True !include non-spherical contributions

KPOINTS:
pymatgen with grid density = 1634 / number of atoms
0
Gamma
4 4 4

POSCAR:
Cr6 N3
1.0
9.5540828705 0.0000000000 0.0000000000
-4.7770425370 8.2740778396 0.0000000000
0.0000000000 0.0000000000 8.8008632660
Cr N
48 24
Direct
0.000000000 0.337082535 0.373132020
0.000000000 0.337082535 0.873131990
0.000000000 0.837082505 0.373132020
0.000000000 0.837082505 0.873131990
0.500000000 0.337082535 0.373132020
0.500000000 0.337082535 0.873131990
0.500000000 0.837082505 0.373132020
0.500000000 0.837082505 0.873131990
0.337082535 0.000000000 0.373132020
0.337082535 0.000000000 0.873131990
0.337082535 0.500000000 0.373132020
0.337082535 0.500000000 0.873131990
0.837082505 0.000000000 0.373132020
0.837082505 0.000000000 0.873131990
0.837082505 0.500000000 0.373132020
0.837082505 0.500000000 0.873131990
0.162917480 0.162917480 0.373132020
0.162917480 0.162917480 0.873131990
0.162917480 0.662917495 0.373132020
0.162917480 0.662917495 0.873131990
0.662917495 0.162917480 0.373132020
0.662917495 0.162917480 0.873131990
0.662917495 0.662917495 0.373132020
0.662917495 0.662917495 0.873131990
0.337082535 0.337082535 0.126867965
0.337082535 0.337082535 0.626867950
0.337082535 0.837082505 0.126867965
0.337082535 0.837082505 0.626867950
0.837082505 0.337082535 0.126867965
0.837082505 0.337082535 0.626867950
0.837082505 0.837082505 0.126867965
0.837082505 0.837082505 0.626867950
0.000000000 0.162917480 0.126867965
0.000000000 0.162917480 0.626867950
0.000000000 0.662917495 0.126867965
0.000000000 0.662917495 0.626867950
0.500000000 0.162917480 0.126867965
0.500000000 0.162917480 0.626867950
0.500000000 0.662917495 0.126867965
0.500000000 0.662917495 0.626867950
0.162917480 0.000000000 0.126867965
0.162917480 0.000000000 0.626867950
0.162917480 0.500000000 0.126867965
0.162917480 0.500000000 0.626867950
0.662917495 0.000000000 0.126867965
0.662917495 0.000000000 0.626867950
0.662917495 0.500000000 0.126867965
0.662917495 0.500000000 0.626867950
0.000000000 0.000000000 0.000000000
0.000000000 0.000000000 0.500000000
0.000000000 0.500000000 0.000000000
0.000000000 0.500000000 0.500000000
0.500000000 0.000000000 0.000000000
0.500000000 0.000000000 0.500000000
0.500000000 0.500000000 0.000000000
0.500000000 0.500000000 0.500000000
0.333333343 0.166666672 0.250000000
0.333333343 0.166666672 0.750000000
0.333333343 0.666666687 0.250000000
0.333333343 0.666666687 0.750000000
0.833333373 0.166666672 0.250000000
0.833333373 0.166666672 0.750000000
0.833333373 0.666666687 0.250000000
0.833333373 0.666666687 0.750000000
0.166666672 0.333333343 0.250000000
0.166666672 0.333333343 0.750000000
0.166666672 0.833333373 0.250000000
0.166666672 0.833333373 0.750000000
0.666666687 0.333333343 0.250000000
0.666666687 0.333333343 0.750000000
0.666666687 0.833333373 0.250000000
0.666666687 0.833333373 0.750000000

Please let me know if I have missed something.

Regards,
Barsha

[martin.schlipf]

When does the error appear? I could go past the parsing of the KPOINTS file in my run. Do I need to run it on a specific number of cores for this error to occur? Do I need to use specific PAWs?

[barshab]

Dear Martin,

The error usually occurs after the number of DOFs has been decided in VASP. So after the following lines of code:

Code: Select all

 Found     6 degrees of freedom:
 Adding 6 more for cell shape distortion
 Finite differences POTIM= 0.01500 DOF=  12

I also see that above I have shown the INCAR for the relaxation. My apologies. I have attached a .zip folder with the correct INCARS and the output and the error from the slurm workload manager can be found.

Best,
Barsha

[barshab]

Dear Martin,

It might also be worth mentioning that I use the Cr_pv POTCAR and the N POTCAR.

Kind regards,
Barsha

[martin.schlipf]

I can reproduce the issue and need to investigate further.

[martin.schlipf]

We will work on fixing this issue as soon as possible. For now, you can set KBLOWUP=F in the INCAR file to change the generation of k points and avoid this issue.

One more point: I noticed that you are calculating the supercell. This issue is already present in the primitive cell. Therefore as a general advice: Always start with the smallest problem size possible and then work up to the actual problem that you want to study. When you work on the smallest problem, you may be able to run this calculation on your desktop or laptop and avoid waiting in the queue only to notice that it is not working. It also allows you to try a few different settings very quickly to figure out what works.

[barshab]

Thank you so much for the advice
I agree, actually, I was thinking the opposite (which is out of inexperience and naivety). When I tried it with the unit-cell (9 atoms), I thought the small number of atoms was the reason that the stress and strain could not fully describe the system and therefore, I tried it on a supercell, thinking it would solve the problem (on having more number of atoms available).
Looking forward to hearing more about it