Temperature going below fix nvt target temperature

LAMMPS LAMMPS General Discussion
1

Hi folks, I was trying to simulate an ionic liquid. I wanted to do simulated annealing with nvt for that I generated velocity at 298 K with random seed and then gave nvt command with Tstart and Ttarget as both 298K for equilibrating but the temperature going to 200K. I am not able to debug why it is happening , If I have packed more amount of molecules then the energy of system should increase by repulsion right? Do someone have any other perspective on this?
I am attaching the datafile and input script below.
embf.in (7.7 KB)
embfnew.dat (5.8 MB)

2

Hi, You have to provide the commands you are using, as well as more information about your system, because there may be several reasons for what you observe. You should have a look at the guidelines of the forum also.
SImon

3

Have uploaded the input script and datafile. Thanks

4

I think this is because you are using the “fix nvt” equations of motions for only a part of the system (more specifically the atoms belonging to the “sol” group) while keeping the rest of the stuff immobile. But when LAMMPS outputs the information on temperature in the log file, it outputs the temperature of all the atoms in the system by default. For you to output only the temperature that concerns the kinetic energy of the atoms in the sol group, you need to use the thermo_modify command.

EDIT: alternatively you can specify in thermo_style command the ID of the compute command you defined to compute the temperature of the sol group. This allows you to output the two things: the temperature calculated considering all the atoms in the system and the temperature computed considering only the atoms in the “sol” group. But if you simply specify the compute ID in the thermo_style, all other thermodynamic properties would still be calculated taking into account all atoms in the system.

1 Like
5

Thanks for your suggestion.