about fix temp/partial

LAMMPS LAMMPS Mailing List Mirror
1

hello,everyone,
this is Rufer.I am modeling friction about a silicon substrate and a silicon tip. the silicon substrate has three layer,fixed layer,newton layer and thermostat layer ,when I give a velocity to tip at a angle relative to X axis,I use fix langevain to control the temperature of thermostat layer .but I need to exclude the impact of velocity in x,y,and load in z,so I use compute temp/partial 0 0 0,but the temperature keeps rising.how should I deal with it, can I change the X axis along the velocity fix it?because compute temp/partial 0 1 0 works well.or use fix nvt?any advice will be appreciated.

Kind regards
Rufer

2

I’m not clear on what you post is asking.
The first thing to do is separate measuring
the temperature that you want from controlling
the temperature. For either, using temp/partial 0 0 0
doesn’t make sense. This is excluding all dimensions
from the temperature, so the result should be 0.0.

This section of the manual discusses how to use
the various temperature computes and how to
also use them for thermostatting in different ways:

https://lammps.sandia.gov/doc/Howto_thermostat.html

Steve