Hi guys.
I’m trying to create a data script for TIP4P water molecule to be used in LAMMPS. But what I have are psf, pdb, topology and parameter files from NAMD. I tried using these files in VMD using the “Add solvation box” tool to generate the water box. Then, in order to generate the data script I need for LAMMPS, I used the “writelammpsdata” command in TopoTools. Then I easily got the complete data script, which includes 2 H atoms, 1 O atom, and the massless atom. However, I read in one of the threads in the LAMMPS Mailing List Archives that explicitly including the 4th site would cause errors in calculating accelerations. Hence, I want to ask:
- What other specific errors would I encounter if I use this data script (having 4 sites) with the LAMMPS pair styles containing /tip4p?
- If it is still possible to use this data script, what should I do to manage the calculation errors?
- Do you recommend that I just use a new data script that contains just H and O (without the massless site) with the /tip4p pair styles?
Thank you so much.
Sincerely,
Hannah
Hi guys.
hi hannah,
I'm trying to create a data script for TIP4P water molecule to be used in
LAMMPS. But what I have are psf, pdb, topology and parameter files from
NAMD. I tried using these files in VMD using the "Add solvation box" tool
to generate the water box. Then, in order to generate the data script I
need for LAMMPS, I used the "writelammpsdata" command in TopoTools. Then I
easily got the complete data script, which includes 2 H atoms, 1 O atom,
and the massless atom. However, I read in one of the threads in the LAMMPS
Mailing List Archives that explicitly including the 4th site would cause
errors in calculating accelerations.
you *can* use the data file, but you cannot use it with fix shake, but have
to use
fix rigid instead, which in turn requires the use of a shorter time step.
Hence, I want to ask:
1) What other specific errors would I encounter if I use this data script
(having 4 sites) with the LAMMPS pair styles containing /tip4p?
the /tip4p styles are designed deal with systems without the M point.
its position is inferred on the fly from the O and H sites. you cannot
use your data file.
2) If it is still possible to use this data script, what should I do to
manage the calculation errors?
see above. you have to use a regular lj/cut/coul/xxx pair style and
keep the water molecules rigid via one of the fix rigid integrators.
3) Do you recommend that I just use a new data script that contains just H
and O (without the massless site) with the /tip4p pair styles?
absolutely.
ciao,
axel.