I want to calculate the mass density profile of a pure organic solution over the course of an npt simulation in LAMMPS. Looking at the documentation there are a number of ways you can do this depending on how you carve up your system. My initial though was simply the mass of the molecules in the simulation (Nmolecules * (molar mass/ Avagadro's constant)) divided by the cell volume. However, if there is a choice in the way you carve up the system surely this will change the volume you are using?
The decision on how to 'chunk' your space can be done by specifying an atom type, a molecule or spatially (1D, 2D or 3D), presumably all these methods would give different volumes so affect the outcome. What would be the most reasonable way?
I think your initial thought seems correct, and the shape of simulation box will depend on your boundary conditions. During the course of NPT simulations the dimensions of simulation box (side lengths and/or angles) will vary and so does the volume, but the shape varies depending on the type of NPT method you use only (for example, shape varying NPT methods do exist.) Thus, the mass density fluctuates during NPT simulations. Another thing, how do you restrict the particles not to leave the simulation box? Do you use molecule or atom based periodic boundary conditions to fold back molecules leaving the simulation box.
Thanks Hiqmet. So there is a thermo style in lammps which allows the mass density of the entire cell to be calculated automatically.
thermo_style density
This is sufficient for my current problem.
I would still be interested to know though how one would best select the appropriate volume if a chunking method was used. I believe the PBC are atomic in this case, but would need to check more thoroughly to be certain.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Atomic, Molecular and Optical Physics
- Atoms