[gmx-users] how to extend the order of Ryckaert-Bellemans dihedral

[Makoto Yoneya]

Dear Erik:

>If we were to implement a special type for it, are there any practical
>limits to the number of terms, and is the format the same as
>ryckaert-bellemans, or?

I saw the expansion up to 12th for the torsion potential of the fluorocarbon
chain in Glaser et al. Liq. Cryst. vol.29(8), '02, 1073-1085, and personally
I need this order (upto 12th).
Since the various potential forms (e.g. sum_n c_n*cos (n*phi) )can be
transform
to the R-B form: sum_n c_n*(cos phi)^n, this R-B form seems to be enough.

>It is actually close to trivial to implement it in the code, the only
>problematic part is that we might need to extend the format in the
>forcefield and run input file.

To quick solve, my specific coefficients could be hand-coded to the GROMACS
source.
Which routine should I change (I could not find these)?

Thank you for advance.

Makoto Yoneya
Yokoyama Nano-structured LC project
Tsukuba, Japan