[lammps-users] temp/rescale

LAMMPS LAMMPS Mailing List Mirror
1

steve,
fix 4 high temp/rescale 1 303.0 303.0 0.1 1.0
compute 1 high temp
fix 3 low temp/rescale 1 297.0 297.0 0.1 1.0
compute 2 low temp

I used compute temp instead of compute temp/region and it’s working fine. The temperature values are matching well now. In my simulaiton, i am first initalising my system using NVT at 300K and then shift to NVE and apply scaling to two plates on both ends and setting up a temp gradient and measure heat current. When i set it to say say 290K and 310K resp’ly the cumilative heat current(change in KE) should be postive as the temp has been increased from 300 to 310K for one plate and should be negative for other plate where T=290. But i observe the opposite trend from the output.

Also on a different note, for a nonlinear lattice where no of atoms in both plates are different, the heat current values are like +ve for high temp side and -ve for low temp side for low temp differences say 295 & 305). But for high temp diff like 270K & 330K on both ends, the heat current values are -ve for high temp and +ve for low temp (opposite of earlier case). What does that mean?? It’s really confusing me…

Thanks
Ajit

2

Also on a different note, for a nonlinear lattice where no of atoms in both
plates are different, the heat current values are like +ve for high temp
side and -ve for low temp side for low temp differences say 295 & 305). But
for high temp diff like 270K & 330K on both ends, the heat current values
are -ve for high temp and +ve for low temp (opposite of earlier case). What
does that mean?? It's really confusing me......

I don't know. The doc page is clear about what value the fix stores. You'll
have to interpret what it means for your model. Maybe the cummulative
aspect of the value is confusing you.

Steve