Questions about the scalar output of NVT

LAMMPS LAMMPS Mailing List Mirror
1

Dear all,

I am using non-equilibrium MD to calculate the thermal conductivity Bi2Te3. I used nvt to fix the temperature at 350K and 250K of the two ends of z direction. The box is periodic in x and y direction. The nvt will return an scalar output that is the cumulative energy change due to the nvt fix. Could I get the heat flux by dividing the scalar output by the simulation time and area? I am also using compute heat/flux to calculate the heat flux. The values in x and y direction are non-zero, and it is difficult to converge in z direction. I first equilibrated the system for 2 ns before sampling 2 ns.

Thank you for any comment and suggestions.

Regards,
Liang Chen

2

I presume you mean you are using two fix nvt commands, one
for each end. You can indeed estimate heat flux with any
thermostat (not just nvt) by monitoring the delta energy in/out
at both ends. I doubt that it will be as accurate as the compute
heat/flux or fix thermal/conductivity commands. See section
4.20 of the manual for more discussion.

Steve