[lammps-users] compute heat/flux

LAMMPS LAMMPS Mailing List Mirror
1

Hello all,
   I am trying to compute the conductivity of Ar-Cu nanofluid with Ar as the base fluid.When i increase the volume fraction of Cu,the obtained conductivity sometimes decreases,which is supposed to increase.Then i visualize the Ar-Cu nanofluid using VMD,i find that the Ar atoms have entered inside the Cu nanoparticles,which is impossible physically.So is there any way to stop the Ar atoms entering inside the Cu nanoparticles?

   Also,the Cu nanoparticles should move as a rigid body during the simulation,i only use EAM potentials to compute the Cu-Cu pairwise interactions without any extra commands to the group atoms of Cu.I am worried that the Cu will no long be a solid particles,that is to say some Cu atoms may move randomly,the conductivity calculated finally may not be correct any more.Do you have any good ideas?(I have tried the fix rigid, fix poems commands but get no good results).

Thank you!
Best regards!
weijie

2

If the particles are doing something unphysical (e.g. Ar going into a
Cu nanoparticle),
then you need to set the Ar-Cu interactions to prevent that. You're creating
the model, so that's up to you. Presumably you have some LJ-like interaction
for Ar-Cu - so try boosting the sigma or epsilon to prevent that from happening.

If you want the nanoparticles to be rigid bodies (as opposed to flexible), then
you can use fix rigid on them.

Steve