[lammps-users] a simple question

By using this fix, I can not get the satifactory results. The temperature of
the system decreases to a very small value which is not the target
temperature . Do you have any tips to get the appropriate damping
parameters of temperature and pressure?

It takes me quite some time, until the temperature control works well.
I use 1000000 Steps just to equilibrate a 2000 atoms metal cell. You
probably should monitor the cell volume and the temperature and let it
run, until a steady state is reached.

Regards, Tobias

If your system doesn't equilibrate rapidly to the thermostat value
then either your damping constant is bad or something else is wrong.
E.g. you are not measuring the same T you are thermostatting.

Unless you have a very non-equilibrium system or start far
from your desired T, it should
not take 1M steps to equilibrate. More like 1000.


Maybe he's making the same mistake I did with the nvt
fix: treating the time units of the damping parameter
as integral time steps rather than the time units
(e.g. picoseconds or femtoseconds determined by the
"units" command).

Yes i did, thank you for the hint. I just copied and pasted it from
the example in the documentation, which is obviously in real units,
all time constants have to be divided by 1000 for metal units.