Thermal conductivity calculation of Carbon Nanotubes using Green Kubo formalism

LAMMPS LAMMPS Mailing List Mirror
1

Dear Lammps users,

I have been trying to find out the thermal conductivity of (11,11)Carbon Nanotube using Equilibrium Green Kubo Relations.
But unfortunately, I have repeatedly been getting values of thermal conductivity which are way off the experimental results.
The values of thermal conductivity that I am getting are around 1.2.W/mk after conversion from Metal to SI units but it should be around 1700 W/mK

System has 528 atoms
Potential : Tersoff Potential

Please help me identify the error that I might be making.

Please find below the lammps input script.

This is the input script for calculating the thermal conductivity of Carbon Nanotube using EGK

echo both
log 2.1_Initialize.out

2

This has been asked on the list before, so you should have a search through the archives to see what's been suggested in the past.

These are the usual suspects:

Is your correlation length long enough for the heat-current autocorrelation function (HCACF) to decay to zero?
Is your simulation long enough to give good statistics in the HCACF?
Is your sampling interval/timestep short enough to resolve the HCACF properly?

You're also calculating a heat flux for the whole system, not just for your nanotube- you should be sure that the value you're comparing to is for the same sort of situation.