Dear All

I'm trying to apply CSHFF to calcium-silicate-hydrate (C-S-H), CSHFF is a re-parameterized version of ClayFF. I built the C-S-H with the coordination of atoms and cell information given by literature, then I built the bond by using "calculate close contacts". There were some defects, but I fixed them manually, and made it similar to the structure pictured in the literature. I also modified the ClayFF based on the parameters given by literature. There was one type of atom, the author didn't give the typing rule, so I gave it a new name and typed it as:

type: cw

  ! interlayer Ca

  template: [>*]

end_type

I also put the new "atom" into all other parts, such as "equivalence_coulombic", if necessary.

Then I tried to run MD with CSHFF with this new force field, however, the structure became disordered. I tried to use other force field, COMPASS worked fine, but COMPASS 2 also made the structure very disordered. So I'm wondering is there anything wrong with the new force field or the C-S-H structure? Actually, the structure I built, if I viewed it from another direction, the bonds looked like prolonged. I checked the lengths, they were about 1.6 angstrom, I think it was reasonable. However,  the COMPASS simulated structure looked fine in every direction with bonds of 1.6 angstrom.

Any one have suggestions about these problems? Thank you.

Regards,

Jingjing