[CP2K-user] QMMM md NVE error GEOMETRY WRONG or EMAX_SPLINE TOO SMALL

[Marcella Iannuzzi]

Dear Antonis

It might be due to how you treat the periodicity of the QM part. 
I noticed that in the QMMM subsection the cell is much smaller, the 
periodicity is set xyz, but the multiples for de-coupling and recoupling 
are set to off.
This looks like a strange choice of settings.
Did you try a to run MM and then restart from an equilibrated MM run?
The MM cell is rather large with respect to the size of the systems. 
Moreover there are fixed atom constraints that also are pretty unusual when 
simulating water.

Regards
Marcella 



On Thursday, July 15, 2021 at 8:58:09 PM UTC+2 ant... at gmail.com wrote:

> Dear Experts/cp2k users,
>
> I am relatively new to cp2k and I was trying to perform a QMMM simulation 
> of water and I am experiencing this error message after a few steps of my 
> simulation: GEOMETRY WRONG or EMAX_SPLINE TOO SMALL. I created a Fortran 
> program to select the QM atoms and save them in an array in order to copy 
> them properly and place them in MM_INDEX input but maybe I am missing 
> something. I have performed classical MD and AIMD test runs with the same 
> coordinates and it seems that the problem isn't from the topology or 
> geometry. I am kind of stack here! Any suggestions would be greatly 
> appreciated.
>
> Please find attached the input files and the geometry/topology I use
>
> Thank you in advance!
> Best regards,
> Antonis
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210716/f2d4867a/attachment.htm>