Problem with temperature in NVT

LAMMPS LAMMPS Mailing List Mirror
1

Dear Lammps community:

I am using LAMMPS (7 Aug 2019). Here I want to prepare my system at the desired temperature 1000K. Using the following command, the system temperature is oscillated and never reaches 800 K. Anyone has a suggestion on this? Thanks in advance.

Best

2

there is not enough information about the kind of system you are simulating to make any definite statements.

in general, it would be better to use a dissipative thermostat like fix langevin at the beginning of equilibration if your system is far from equilibrium (e.g. needs to be melted). also, in such cases, you can assign a much higher initial temperature, as up to half of the kinetic energy will be converted into potential energy during equilibration.

also, 10000 MD steps is not a lot for equilibrating all but the most trivial systems.

axel.

3

Thanks very much for your reply.

I want to prepare the Fe surface slab at the desired 1000 K. Firstly, I minimize the system. Then I used FIX NVT for 1000 K sampling. Since the “DRAG” option may change the distribution of velocity and position, I use firstly 2000 steps for FIX NVT with DRAG, followed by a FIX NVT without DRAG for another 1000 steps. I hope this way the velocity and position are correct, namely, corresponding to 1000 K. Is this way suitable?

Thanks. See also below for detailed input commands.