problem in fixing the outer tube with heat transfer or other effects only from/to the inner tube

dear all,
I have a problem which confused me for long time when I simulate the relatie sliding of walls in a Double-walled carbon nanotube. I fix the inner tube in NVT ensemble with 300K and move the outer tube partially. My command like this:

fix 1 inner nvt temp 300 300 0.1
fix_modify 1 energy yes
fix 2 part_of_outer move linear 0 0 0.5

Wth the inner tube partially fixed, I want to know whicn ensemble or what commands I should use to the other part of the outer tube, to keep the temperature of the outer tube close to the inner tube. At the same time, I don’t want another NVT to exchange energy with outer tube and I also found “fix 3 outer_part nvt temp 300 300 0.1” doesn’t work as I wish, the temperature get stable at far away for 300K(250K for translation motion and 600K for rotation). I just hope the outer tube could be near 300K only with the inner tube affect.
Do you think this possible? I get struck in this problem for dozens of days. Any help will be appreciated. Thanks very much!

Yours sincerely,
Ma Teng

If you are saying that part of the outer tube is not
being moved directly via the fix move command
and you want to thermostat those atoms …

You can do it with any of the thermostats in LAMMPS,
e.g. fix langevin, fix nvt, etc. Note that since those
atoms are part of a CNT they are probably moving with
the ones that are sliding via fix move. So the “thermal”
temp of the other atoms in the CNT needs to have the

linear motion subtracted off. See Section 6.16 of the manual
for how to compute a temperature that applies such

a bias, and how to use that temperature with a thermostat.

Steve