Questions about Commands

Dear all,

I have questions about commands of GULP:

  1. I want to use the fifth-order Ryckaert-Bellemans torsion potential. What are the meanings of ‘<flagsnorder+1>’ and 'nordercoefficients’ in the help file? Can you give me an example how to express the fifth-order Ryckaert-Bellemans potential?
  2. I found some examples. What are the meanings of kcal in the command ‘lennard 12 6 zero kcal epsilon combine product all’ and ‘harmonic intra bond single kcal’?
  3. How can I set the coulombic 1-4 scale factor? I used the command ‘coulomb_subtract x13 0.5’, but when I output lammps file, I found the coulombic 1-4 scale factor in special_bonds is 0. And I used the command’lennard epsilon geometric 12 6 x13 all 0.5
    10.0’, and the lj 1-4 scale factor in lammps file is 0.5, which is normal.

Best Regards,
Wang Qiang.

Hi Wang Qiang
Here are some quick answers:

  1. If you are using a fifth order polynomial then there are 6 parameters and so the help means that you supply 6 fitting flags (if fitting) and “norder”, i.e. 5 coefficients on the next line since the overall scaling parameter k0 is given on the previous line (making the 6 in total).

  2. Many of these are explain in the manual in the Table on page 108. The remaining ones are given in the help text for the potential type. For example, zero for lennard:

     : Specifying "zero" allows the user to chose between sigma defined
     : as the potential energy minimum distance (default) or the distance
     : at which the potential energy goes to zero.
  3. You’ve set the Coulomb scale factor - it’s the half on the input line you quote. It’s important to note that GULP isn’t designed as a tool to create LAMMPS input files & so there are no guarantees that it will set things up 100% correctly. If you want to run LAMMPS it’s better to directly create the inputs - the capability in GULP is just to give you a starting point for translating inputs.


Dear julian,

Thanks for your reply.

I have performed MD simulations using Lammps. And now I want to perform lattice dynamics using GULP. I need to make sure that the force field used in the two software is correct and the same.

The formula of Ryckaert-Bellemans four-body potential is

In my case, the four atoms are C1,C2, C3 and O1. And I have obtain k0,k1,k2, k3 and k4 (fifth order). I am still confused about what you said and the parameters required by the help file:

Can you give me more clues? Thanks again.

Best Regards,
Wang Qiang

The input would look something like:
C1 C2 C3 O1 k0 1.5 1.5 1.5 4.5
k1 k2 k3 k4 k5 k6

where k0-k6 are the coefficients for the n th power of cos phi.
The maximum distances between atoms (rmax) are guesses for your cutoffs for C-C/C-O bonds with the 1-4 distance set to be the sum of the 3 bonds to be safe.
Hope that helps,

Dear julian,

Thanks for your help!

Best Regards
Wang Qiang