I am trying to find how the dataset in this article was generated. There is a description in the article on the methods used but I’m not a skilled LAMMPS user.
The system is a simulation of 4096 atomic particles (two kinds of particles) in a cube of side length L with periodic boundary conditions with typical Kob Andersen configuration parameters. The articles says:
“To obtain an equilibrated system at the desired temperature and pressure, we perform several stages of simulations to safely reach a desired state point. First, we equilibrate the system at a starting high-temperature state point (TS = 0.5495, PS = 0.4888) using an isothermal-isobaric ensemble. We then perform a gradual quench to the desired state point (T_target, P_target) using the same method, followed by one final run at the target state point to help preserve the pressure and temperature. We then switch to a canonical ensemble to further stabilize the temperature and pressure and obtain an equilibrated system. We run each part of this equilibration process for at least ten relaxation times.”
I am starting the simulation with random particles whose velocties are distributed like a Maxwell Boltzmann with the command
region my_region block 0 9.4 0 9.4 0 9.4 create_box 2 my_region create_atoms 1 random 3277 229609 my_region create_atoms 2 random 819 691203 my_region mass 1 1 mass 2 1 velocity all create 0.5495 92561 dist gaussian
Is it right to define the box volume for the starting configuration if I first run an npt (Isothermal Isobaric ensamble) in which the volume can change?
Why the article says: “We simulate a system of N = 4,096 particles in a cube of side length L with periodic boundary conditions at various temperatures and pressures.” if the volume during the npt simulations can change? Am I missing something?
- To run the npt simulation I’m using the command
fix nose all npt temp 0.5495 0.5495 $(100.0*dt) iso 0.4888 0.4888 $(100.0*dt)
I’m using Lennard Jones units, i don’t know if that is right.
I I understand correctly, the simulation in the article were performed by running a series of npt simulation at different temperatures and then an nvt (canonical ensemble). How can i feed the final data from a simulation to another? Should I use the write_restart command?
Is there a way to estimate the relaxion time of the systems?
Can I post this same message to the LAMMPS mailing list?