Start noncollinear calculation from WAVECAR of collinear calculation

[johan_felisaz]

Hi,

I would like to start a noncollinear calculation, from the WAVECAR of a previous collinear calculation.

Is it even possible ?
Whenever I put ISTART=1 in my INCAR file, I get the following error:

Code: Select all

 -----------------------------------------------------------------------------
|                                                                             |
|     EEEEEEE  RRRRRR   RRRRRR   OOOOOOO  RRRRRR      ###     ###     ###     |
|     E        R     R  R     R  O     O  R     R     ###     ###     ###     |
|     E        R     R  R     R  O     O  R     R     ###     ###     ###     |
|     EEEEE    RRRRRR   RRRRRR   O     O  RRRRRR       #       #       #      |
|     E        R   R    R   R    O     O  R   R                               |
|     E        R    R   R    R   O     O  R    R      ###     ###     ###     |
|     EEEEEEE  R     R  R     R  OOOOOOO  R     R     ###     ###     ###     |
|                                                                             |
|     ERROR: while reading WAVECAR, plane wave coefficients changed 27198     |
|     13599                                                                   |
|                                                                             |
|       ---->  I REFUSE TO CONTINUE WITH THIS SICK JOB ... BYE!!! <----       |
|                                                                             |
 -----------------------------------------------------------------------------

I understand why it complains (because goin to spinors doubles the number of coefficients), but VASP should be able to just copy them right ?
Furthermore, the wiki ( wiki/index.php/LNONCOLLINEAR ) and previous forum answers ( forum/viewtopic.php?p=18748#p18748 ) hints that it should be possible.

For reference, here is my INCAR file:

Code: Select all

lnoncollinear = true
nbands = 48
ispin = 2
system = new_seq
magmom = 0 0 0 0 0 0 0 0 0
lasph = true
kpar = 1
encut = 600
gga_compat = false
sigma = 0.0025
lmaxmix = 4
istart = 1
icharg = 1

(I did not add my other input files because the exact details of my system don't matter)

Thanks in advance,

[pedro_melo]

Dear johan_felisaz,

Did you follow the steps detailed in the answer to the post you linked? Was vasp_ncl used in all of them? While the question is general, your input and output files can help us track if there is something specific with your calculation, so please attach them.

Kind regards,
Pedro

[johan_felisaz]

Hi,

Yes I followed the steps closely. vasp_ncl was used all along.

You will find attached my files.

[pedro_melo]

Dear johan_felisaz,

You are using the same number of bands in both calculations. When running the noncollinear calculation you should double the number of bands that you used in your previous step.

Also, please note that your POTCAR file has the pseudopotentials in the wrong order. Right now you have the pseudo for Nb and then for Se, it should be the other way around. The order should be consistent with the information in your POSCAR file.

Kind regards,
Pedro

[johan_felisaz]

Hi,

Indeed, it was as simple as that. It works now, thanks !