running external fortran program for energy and force calculations with rigid body dyanmics

[Ommair Ishaque]

Hi,

I want to run classical MD with rigid body dynamics for I have locally written program for water potential (calculate energies and forces). Can anyone guide me is it possible to run classical MD simulation with minimum 256 water molecules in a box as rigid body. I would like add only no-bonded interactions such as repulsion, dispersion, Coulombic  and induction (dipole-dipole interaction) which already in my code.

If yes can anyone give some example similar to this

best 

Ommair   

[Michele Ceriotti]

You can use driver.f90 as an example of how to interface your code with i-PI, and examples/lammps/h2o-constraints as an example of how to set constraints. 

[Ommair Ishaque]

And can I also use massless sites in water molecule? Is it possible in i-pi to handle rigid body dynamics with massless sites?

[Michele Ceriotti]

No, not easily. You need to take care of all the unphysical stuff on the client side. Should actually be easy to implement a TIP4P-like model without explicitly treating the M sites as atoms. LAMMPS does that. Actually even driver.f90 implements a 4-sites water model without having explicit atoms for the centers of charge.