NEB replicas?

220 views
Skip to first unread message

Ernst Larsson

unread,
Mar 26, 2021, 7:35:03 AM3/26/21
to cp2k
Dear all,

I'm slightly confused about the "NUMBER_OF_REPLICA" keyword when performing a NEB-calculations.
I've optimized the start and end points separately, which due to symmetry have the same energies.

When I run the NEB-calculation, I provide three intermediate starting guesses which I've interpolated myself (I know that CP2K in principle can do this for me) and set "NUMBER_OF_REPLICA = 5".

My confusion comes when I look at the energies and files:
1) I get six files of the type '...-BANDX.out' rather than five.
2) The ENERGIES section in the NEB-output give me five energies, which is fine, but only the start point matches what I got from the geometry optimisation. The last replica has an energy which is larger than what it should be by 0.011205 a.u., (when it should be identical to the first). If I check "...-pos-Replica_nr_5-1.xyz" which I presume is for the endpoint, it is not identical to what I specified as my end point in the input.
3) The energy in this extra '...-BANDX.out'-file does not match the correct end-point either.

Does anyone have any good idea on what may be going wrong?
I've attached my NEB-input script.

Regards,
Ernst
MgO_100.inp

fabia...@gmail.com

unread,
Mar 26, 2021, 8:01:21 AM3/26/21
to cp2k
Hi Ernst,

You can set ALIGN_FRAMES and  ROTATE_FRAME to .FALSE. to prevent this behavior.

Cheers,
Fabian

Ernst Larsson

unread,
Mar 26, 2021, 8:37:01 AM3/26/21
to cp...@googlegroups.com
Dear Fabia,

Thank you very much!
The additional keywords seem to have solve the energy issues.
I still get one extra ’…-BANDX.out”-file though….

Regards,
Ernst

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/f62df28c-6bed-4a3e-8454-4ee62d645836n%40googlegroups.com.

Marcella Iannuzzi

unread,
Mar 27, 2021, 7:22:19 AM3/27/21
to cp2k
Dear Ernst, 

The extra output file should be due to the ratio between the the total number of tasks and the number of tasks per replica, 
which I suspect does not correspond to the number of replicas. 
Best
Marcella
Reply all
Reply to author
Forward
0 new messages