Switching from LAMMPS/DL_POLY to CP2K: Avoiding multiple NONBONDED declaration

[kacper.d...@gmail.com]

Dear All, 

I am trying to transfer the Amber-type force field into CP2K.

While translating bonds, angles, and dihedrals definitions were straightforward I feel confused in defining the nonbonding pair interactions.

My nonbonding interactions are defined in terms of both Lennard-Jones and Buckingham definitions.  See the file attached (the parameters in the attachment are written according to the format of the FIELD file of the DL POLY code).

I would like to use both Lennard-Jones and Williams's potential (I assume that Williams and Buckingham's potential definitions are very similar if not identical.)

I assume that I can either express the NONBONDED terms in the main CP2K input (&nonbonded along with explicit definitions of charges, bonds, etc.) or in the external potential file, but the external file only allows NONBONDED definitions within Lennard-Jones scheme? Am I correct?

Please forgive me, but I am very much confused and any error at this stage will get me completely lost. Any suggestions?

With best wishes,

Casper

[Krack Matthias (PSI)]

Hi Casper

 

The &NONBONDED subsection &WILLIAMS defines indeed (exactly) a Buckingham potential. I prefer to define the nonbonded interactions in the CP2K input. You can include the &NONBONDED section from an external file using @INCLUDE to keep the main CP2K input file short.

Pay attention to the correct conversion of the potential parameter units. CP2K provides various units which can be combined like in this test input.

 

HTH

 

Matthias

--
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/f8ffd47f-eef5-4ef7-b428-155d7a167f84n%40googlegroups.com.

[kacper.d...@gmail.com]

Dear Matthias, 

Million thanks for your kind help. I can confirm that eventually everything works perfect to me. I was able to perfectly translate the force field definitions from DLPOLY into CP2K for a very lengthy case (luckily of AMBER type). Indeed,  I would like to bring to your attention importance of the force field units and slight differences between the Buckingham and Williams definitions. For the ones translating the potentials from DL_POLY *.Field files, I would advise you to double check the inputs referring to:

https://www.chemshell.org/sites/www.chemshell.org/files/docs/tcl-chemshell/manual/dl_poly_print.html

The best way to me is to define everything in the input file, with the units defined explicite for each line.

Having an opportunity to use both classical force-field MD and ab initio MD using the same code is absolutely amazing! Thank you!

All the best,

Casper