I am beginner in lammps. I would like to model a hexagonal close-packed (HCP) lattice (a=3.23, c=5.23) with gas molecules placed randomly in the octahedral or tetrahedral interstitial sites of the lattice. Could you please guide me on how to implement this in LAMMPS?
Also, How can I ensure that gas atoms are only placed in the interstitial (octahedral or tetrahedral) voids of the HCP lattice during modeling in LAMMPS?
Manually providing the coordinates for each gas atom is time-consuming. Is there an easier or automated way to place gas atoms only in interstitial voids of the HCP lattice in LAMMPS?
You could define a custom lattice that has all possible positions for the gas molecules within the unit cell. Then you can create a molecule file and read it with the molecule command. Then you can create the molecules at the desired locations at the desired fraction of locations. Finally you change the lattice to the regular HCP lattice and create the atoms for that lattice.
You will need to do some minimization, possible with replacing the regular force field with the soft pair style to reduce overlaps first before continuing with the real force field. And then continue as normal with equilibrating this system at the desired conditions.
Altogether, as a beginner in LAMMPS you are recommended to practice individual parts first by working though several related tutorials and only when you are confident to do every step correctly, should you attempt to build your system for your project. Skipping ahead will result in many problems due to mistakes where you don’t know whether you did something wrong or whether the system needs to more care and special treatment to be guided toward the desired final configuration.
That would be your job to check that: If you know how to mathematically write the criteria, you can print out the positions of the atoms and perform the analysis with a custom script. It should also be possible to implement it into a LAMMPS input script using some compute commands.
No, that is not what I suggested. I suggested you first create a custom lattice with the desired positions and then have a random partial occupancy, not random positions.
Of course not, but that is because you put them into random positions and not randomly occupied a subset of the desired positions.