The initial seperation distance between layers is 2nm, and I want to calculate the thermal conductivity along x-axis. So my assumption is after the equilibration the seperation between layers should not shrink but same as initial position
The initial seperation distance between layers is 2nm, and I want to calculate the thermal conductivity along x-axis. So my assumption is after the equilibration the seperation between layers should not shrink but same as initial position
well, since there are no interactions in z-direction (2nm is _far_
beyond the cutoff), there
will be nothing stopping the layers from drifting, once they start due
to some numerical
noise. you can probably alleviate that through applying a z-direction
only fix spring with
a _very_ low spring constant to each layer.
the bigger issue is that you won't be able to capture the (very likely
to be significant)
contribution of the electrons to the thermal conductivity?
axel.
other than by using fix spring, I have a query. So does it means I should seperate my graphene sheet within cutoff?
other than by using fix spring, I have a query. So does it means I should seperate my graphene sheet within cutoff?
once they are beyond the cutoff, it doesn't make a difference how far
they are apart.
with AIREBO and 3sigma cutoff, you have a cutoff of 10.2 angstrom. any atoms
that are farther apart don't interact. so said you spaced them 2nm =
20 angstrom apart,
so they don't interact. with fix spring you'll apply a soft confinement.
if you _do_ want interaction between the layers, then you have to have
them closer,
but then - of course - they will sooner or later collapse to a stack of sheets.
axel.
One more thing, will as you said the cut off, so will the airebo LJ term and torsion term on/off will affect the graphene simulation?
How could it not?
Steve
Dear Jun Kit,
Maybe you should check your periodic boundary condition along x-direction or y-direction or both of them.
rizal
2011/3/17 Steve Plimpton <sjplimp@…92…>