Hi, Steve
I used a smaller timestep 0.001ps. And increasing the damping constant (0.1) for langevin. But my results still have temperature jump. See the plot below. The first bin and last bin are the silver wall I maintained temperature using langevin. Other wall bins are free.
Thanks
ziyuan
export.wmf (9.07 KB)
Ziyuan,
What you are seeing is normal because the grouped bin for temperature controlling is pretty small, which naturally raise the temperature jump. If you want to get rid of it, increase the group region for thermostat.
BTW, as far as I know, thermal transport in Silver is electricity driven dominant, how are you gonna do with this only using classic MD simulations?
Best,
Ajing
export.wmf (9.07 KB)
Thanks Ajing. You are right.
ziyuan
export.wmf (9.07 KB)