Dear LAMMPS users,
I tried to measure the thermal conductivity of a bilayer structure. I tried to use NVT to control the temperature of hot/cold reservoir. However, I found that I can’t control them well compared to the target value. I can’t figure out the reason but I guess there might be two potential ones: 1. The system is small and the number of atoms (around 350) in hot/cold reservoirs is not enough. 2. Tdump (50) is not proper.
Can any one give me some advices?
Thank you very much
Sincerely,
Yuan Gao
Dear LAMMPS users,
I tried to measure the thermal conductivity of a bilayer structure. I tried
to use NVT to control the temperature of hot/cold reservoir. However, I
found that I can't control them well compared to the target value. I can't
figure out the reason but I guess there might be two potential ones: 1. The
system is small and the number of atoms (around 350) in hot/cold reservoirs
is not enough. 2. Tdump (50) is not proper.
Can any one give me some advices?
fix nvt uses nose-hoover thermostat chains. those are designed to
sample the canonical ensemble for a system in equilibrium. they are
not as well suited for non-equilibrium systems.
350 atoms is indeed a small number and you can expect quite large fluctuations.
it is difficult to give (good) advice on such a minimal amount of information.
all three of these points are frequently discussed and explained on
this list. please study the mailing list archives.
axel.