Issues in Band Renormalization Calculations with VASP 6.5.1

[jiong_yang1]

Hello! I have encountered two problems while using VASP 6.5.1. I look forward to your answers!

1.When calculating the phelel files of some 2D materials, I found that the structure in CONTCAR_ELPH has changes in the coordinate axes and structure compared to the originally optimized POSCAR. This leads to the problem where NGX cannot be equal in subsequent unit cell calculations. May I ask why this structural change occurs? How can this be avoided?
Origin POSCAR:

Code: Select all

"P4"                                    
    1.000000000000000     
     4.6226669632130948    0.0000001381510300    0.0000000000000000
    -0.0000000117919767    3.2984462344155139    0.0000000000000000
     0.0000000000000000    0.0000000000000000   20.0000000000000000
  P             
     4
Direct
     0.0082061201240222    0.9999122400892801    0.4474203012014542
     0.1865431330963787    0.9999122543054995    0.5525796087985455
     0.5082061479036213    0.4999122406944991    0.5525796087985455
     0.6865431048759802    0.4999122539107148    0.4474203012014542
 
  0.00000000E+00  0.00000000E+00  0.00000000E+00
  0.00000000E+00  0.00000000E+00  0.00000000E+00
  0.00000000E+00  0.00000000E+00  0.00000000E+00
  0.00000000E+00  0.00000000E+00  0.00000000E+00

CONTCAR_ELPH：

Code: Select all

unknown system                          
    1.000000000000000     
     0.0000004520287952   -3.2986588638309042    0.0000000000000000
     4.6248232683044250   -0.0000008464783032    0.0000000000000000
     0.0000000000000000    0.0000000000000000   20.0000000000000000
  P             
     4
Direct
     0.0000876853304987    0.0081315433044087    0.4476227604924858
     0.0000878527317738    0.1866177312265300    0.5523771495075067
     0.5000876522682702    0.5081315497740073    0.5523771495075067
     0.5000878206694566    0.6866176816961591    0.4476227604924858

2. To continue the calculation, I selected CONTCAR_ELPH to directly expand the unit cell and performed the phelel calculation again. After the calculation, I used the phelel file and CONTCAR_ELPH (copied to POSCAR) for the next step of band renormalization calculation. At this point, the calculation terminates with an error:

Code: Select all

 -----------------------------------------------------------------------------
|                     _     ____    _    _    _____     _                     |
|                    | |   |  _ \  | |  | |  / ____|   | |                    |
|                    | |   | |_) | | |  | | | |  __    | |                    |
|                    |_|   |  _ <  | |  | | | | |_ |   |_|                    |
|                     _    | |_) | | |__| | | |__| |    _                     |
|                    (_)   |____/   \____/   \_____|   (_)                    |
|                                                                             |
|     internal error in: wave_rotator.F  at line: 248                         |
|                                                                             |
|     Number of g-vectors is incompatible 6582 6580                           |
|                                                                             |
|     If you are not a developer, you should not encounter this problem.      |
|     Please submit a bug report.                                             |
|                                                                             |
 -----------------------------------------------------------------------------

application called MPI_Abort(MPI_COMM_WORLD, 1) - process 18

May I ask what causes this, and how should we modify it?
Relevant files are in the attachment.
Thank you, and I look forward to your response!

[marie-therese.huebsch]

Hello, thank you for your inquiry.

Ad 1)

This may be a result of SYMPREC or the proceedure you used to create a supercell. If you want we can investigate that in more details. To this end please provide the corresponding reproducible including how you create the supercell and the POSCAR of the supercell. If you only want to avoid this in the future, the ELPH_POT_LATTICE tag can be used to specify the primitive lattice vectors.

Ad 2)

We would like to reproduce this bug. Could you please provide the input for all steps of the calculation (also to produce the phelel)? And the corresponding submittion script so I clearly see which files correspond to which job. Please also add the OUTCAR for all steps of the calculation.

Best regards,
Marie-Therese

[jiong_yang1]

Hello! Thank you for your support. Here are my calculation files. I wonder if these are sufficient? It should be noted that there are two Phelel calculation input folders. The first one contains the original input. After inputting, I noticed a rotation occurred, so I performed a second Phelel calculation using the rotated structure. The subsequent unit cell calculations use the second Phelel file. If you have any other needs, please feel free to let me know! Thank you very much!

[marie-therese.huebsch]

Sorry for the long scilence. Unfortunately the forum had a system change and we did not notice that the notification where not send for quite some time. So I totally missed your response. I will look into it this week and get back to you with an update asap.

[manuel_engel1]

Hi,

I'm from the electron-phonon team and I'm currently working through your calculation. Unfortunately, I am unable to reproduce the CONTCAR_ELPH you have shown in the very beginning. When I run your phelel-supercell-1 calculation, it will result in the following primitive unit cell (CONTCAR_ELPH):

Code: Select all

unknown system
   1.0000000000000000
     0.0000000117919767   -3.2984462344157954    0.0000000000000000
     4.6226669632128319    0.0000001381510298    0.0000000000000000
     0.0000000000000000    0.0000000000000000   20.0000000000000000
  P
     4
Direct
     0.0000877599108053    0.0082061201240085    0.4474203012014540
     0.0000877456945859    0.1865431330964035    0.5525796087985454
     0.5000877593055292    0.5082061479038851    0.5525796087985454
     0.5000877460893277    0.6865431048762396    0.4474203012014540

This is consistent with the supercell from your phelel-supercell-1 folder. You can also confirm this in the OUTCAR file by looking at the very end:

Code: Select all

 Generation of phelel_params.hdf5
 ================================
 
Primitive cell
   a1 =     0.00000001   -3.29844623    0.00000000
   a2 =     4.62266696    0.00000014    0.00000000
   a3 =     0.00000000    0.00000000   20.00000000

Could you please detail the steps that you took to obtain this CONTCAR_ELPH? Before I can continue to investigate your second issue regarding the bug message, I need to understand what has led to this point.

[jiong_yang1]

Dear ,
We sincerely apologize for the delay in replying to you due to some special circumstances. As a matter of fact, we did not proceed with the previous calculations, and thus no additional files have been retained.
However, we encountered the same issue while calculating the new system. Hereby, we provide you with all the calculation files except the phelel.hdf5 file, as it is excessively large.
We apologize again for any inconvenience caused and look forward to your further reply. THANKS！

Input Links： https://github.com/Zhaoyihang-PRC/MgCl2-BUG.git

Code: Select all

 -----------------------------------------------------------------------------
|                     _     ____    _    _    _____     _                     |
|                    | |   |  _ \  | |  | |  / ____|   | |                    |
|                    | |   | |_) | | |  | | | |  __    | |                    |
|                    |_|   |  _ <  | |  | | | | |_ |   |_|                    |
|                     _    | |_) | | |__| | | |__| |    _                     |
|                    (_)   |____/   \____/   \_____|   (_)                    |
|                                                                             |
|     internal error in: wave_rotator.F  at line: 248                         |
|                                                                             |
|     Number of g-vectors is incompatible 6289 6291                           |
|                                                                             |
|     If you are not a developer, you should not encounter this problem.      |
|     Please submit a bug report.                                             |
|                                                                             |
 -----------------------------------------------------------------------------

 -----------------------------------------------------------------------------
|                     _     ____    _    _    _____     _                     |
|                    | |   |  _ \  | |  | |  / ____|   | |                    |
|                    | |   | |_) | | |  | | | |  __    | |                    |
|                    |_|   |  _ <  | |  | | | | |_ |   |_|                    |
|                     _    | |_) | | |__| | | |__| |    _                     |
|                    (_)   |____/   \____/   \_____|   (_)                    |
|                                                                             |
|     internal error in: wave_rotator.F  at line: 248                         |
|                                                                             |
|     Number of g-vectors is incompatible 6315 6313                           |
|                                                                             |
|     If you are not a developer, you should not encounter this problem.      |
|     Please submit a bug report.                                             |
|                                                                             |
 -----------------------------------------------------------------------------

application called MPI_Abort(MPI_COMM_WORLD, 1) - process 2

[manuel_engel1]

Hello,

No worries! And thank you for providing all these input/output files. It will be much easier to diagnose the problem with this. Please understand that it will take some time to reproduce your error message and to understand the problem. I will reply to this topic again once I have new information.