[lammps-users] About fix/rescale problem

Hello, this is rufer, I am trying to repeat the model in the article: Electrotunable Lubricity with Ionic Liquid Nanoscale Films. The system has two walls, one is fixed, the other is pressed in z-direction, The walls were kept at a constant temperature of T = 600 K by rescaling the velocities of the walls atoms. In my opinion, the fixed wall can be set by:

fix 1 fixedwall setforce 0 0 0,
and the pressed wall can be set by:
fix 2 pressedwall aveforce NULL NULL v_load,
v_load is the pressure which will be set in direction, the temperature of two walls is set by:
fix 3 wall temp/rescale 10 600 600 0.05 1.0
when I run the simulation code mentioned above,the system will blow up, I don’t know why, all the pair coeffs is set as same as this article, the units is right, so what is the problem? or how to apply fix/rescale in this simulation? Any advice will be appreciated!

Best wishes.

If you use fix setforce 0.0 0.0 0.0 and fix temp/rescale on the same atoms, it will simply disintegrate.
Just consider Newton’s equations of motion. Because of the scaled velocities the positions will change into whatever direction the (random?) initial velocities are pointing.

The description you provide does not make much sense and using fix temp/rescale is never a good choice for real production simulations.

That said, if you need to get advice about how this published simulation was realized the proper people to ask are the (corresponding) authors.
Nobody else knows most of the details. Anybody else can only infer what is in the publication just as well as you could.

Thanks for your advice! Actually, I am using fix nve in my in script, so the velocity and location of the particle will be computed. I will try to ask for the author’s help. Best wishes!

Axel Kohlmeyer <[email protected]> 于2021年11月11日周四 上午11:15写道: