QMMM md NVE error GEOMETRY WRONG or EMAX_SPLINE TOO SMALL

[Antonis Evang]

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

[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