[lammps-users] water,TIP4P

LAMMPS LAMMPS Mailing List Mirror
1

Dear Friends

I want to simulate TIP4P water molecule, but I don’t know how should I define the 4th site of the molecule. Should I define the site as an atom that it’s mass and charge are zero and it’s bond_style with oxygen is harmonic that it’s kb equals a high value and bond lenght equal to OM in Manual?

Thanks alot for your time and attention

Saly

2

Dear Friends

I want to simulate TIP4P water molecule, but I don't know how should I
define the 4th site of the molecule. Should I define the site as an atom
that it's mass and charge are zero and it's bond_style with oxygen is
harmonic that it's kb equals a high value and bond lenght equal to OM in
Manual

have a closer look at the LAMMPS documentation. there are several
pair styles with /tip4p appended that allow reconstruction of the M site
from the O and H sites by specifying the O and H site atom types.
this is more efficient than defining those as rigid molecules (the alternative).

axel.

3

See Section 4.8 of the manual.

Steve

4

I am building my first model of tip4p water with no boundaries, and I have a question about the charge parameter for the Oxygen atom listed on the lammps page for the tip4p model.

On the lammps page it states that the charge for the O atom for tip4p should be -1.04.

However in the reference paper listed on the lammps page. (Jorgensen et al. “Comparison of simple potential functions for simulating liquid water”, Journ. Chem. Phys vol 79, 1983),
in table 1, the charge for Oxygen is listed as 0.0, and the charge for the dummy site M is listed as -1.04, for the tip4p model.

Do I just go with the parameters on the lammps page, and put a charge of 0.0 on the M site?

These contradictory statements are very confusing, as I’m still just starting out at this.

Guidance greatly appreciated,

Hello, Daniel:

You do not create the dummy sites explicitly in LAMMPS, as the creation of a site with zero mass leads to errors in the calculation of accelerations. Instead, you use the parameters in the lj/cut/coul/long/tip4p specification to determine which atom type has the dummy charge. Assigning the charge to that atom, it gets transferred to the dummy atom by the code.

At the current time, however, I am debugging some issues associated with the virial calculations in the code. I hope to get those resolved in the next week or so.
–AEI