"GEOMETRY wrong or EMAX_SPLINE too small!" error in FIST classical MD of water

[Akanksha]

Hello all,

I am trying to run classical MD for a water box on the CP2K 2024.2 version using the input file from this exercise:

https://www.cp2k.org/exercises:2018_ethz_mmm:h2o_md

I kept my input parameters exactly the same. And the calculations went through smoothly.

Next, the only difference I made was - 

I put the given coordinates into a separate .xyz file (attached with the email) and corrected it as needed into the proper chemical file format. Then, I changed the "subsys" section as follows -

  &SUBSYS
    &CELL
      ABC 9.865 9.865 9.865
    &END CELL
    &TOPOLOGY
        COORD_FILE_FORMAT   XYZ
        COORD_FILE_NAME       ${COORDFILE}
    &END TOPOLOGY
  &END SUBSYS

In this case, the calculation stops immediately, and I get the error - 
"GEOMETRY wrong or EMAX_SPLINE too small!"

I have tried these things but I get the same error every time:
1) Used a completely different xyzfile. 
2) Moved some molecules around so they are not that close in the initial geometry
3) Increased the value of EMAX_SPLINE to 2.0

I do not understand why exactly the same coordinates would not work in different file formats when I haven't changed anything else. Could anyone please help me with this?
Thank you very much!

[Jürg Hutter]

Hi

you most likely have 'overlapping' atoms due to PBC. Please check your input coordinates
for this point.

best
JH

________________________________________
From: 'Akanksha' via cp2k <cp...@googlegroups.com>
Sent: Sunday, March 30, 2025 10:44 PM
To: cp2k
Subject: [CP2K:21312] "GEOMETRY wrong or EMAX_SPLINE too small!" error in FIST classical MD of water

--
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<mailto:cp2k+uns...@googlegroups.com>.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/921969ad-fcd6-46b6-a092-dd8907701737n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/921969ad-fcd6-46b6-a092-dd8907701737n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Akanksha]

Hello Prof. Hutter,

Thank you for getting back to me. I checked the periodic images for any overlap, but there wasn't any. However, I realized the error was caused by missing connectivity information. After adding the .psf file, it worked out fine.

Thanks again for your help,

Akanksha

To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/ZR0P278MB07599FDA7B4BA06751DE27B49FAD2%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.