Dear lammps users,
Has anybody modified lammps to be able to use Compute heat/flux (Green-Kubo method) using Tersoff potentials?
I know that lammps currently has Compute heat/flux only for pair-potentials.
Of course, one can use Muller-Plathe method instead. But my experience with Muller-Plathe tells me it is very sensitive to the interference of the simulation size with the mean free path.
Your help is greatly appreciated.
Regards
Babak Kouchmeshky
Civil and Environmental Eng. Dept.
Cornell University
no - but there will be a new comptue heat/flux soon
that will do all potentials except long-range Coulombics.
Steve
Dear Babak,
There is long discussion on the forum how to calculate heat flux for
any potentials using
stress per atom values. You can search through e-mails.
I could not obtain any reasonable value for thermal conductivity in Si
(diamond structure) with Tersoff
potential by any method (Muller-Plathe, Gree-Kubo or by direct heating
and cooling of
two regions), but for SW or MEAM potentials all this methods,
including G-K, works good.
It will be interesting for me to hear about your experience with
Tersoff potential.
Best,
German