Energy conservation of SRD fluid in example output files

LAMMPS LAMMPS Mailing List Mirror
1

Hi Steve,

The total energy of the srd fluid is supposed to be conserved (potential energy is zero for srd) for the examples/srd/in.srd.pure as it is done in NVE ensemble. I checked the example output files in different versions of the lammps and it is conserved only in the older version (15Jan11) but not in the newer versions (8Jan13 and 14Oct12).

Only this example output file looks fine- lammps-15Jan11/examples/srd/log.srd.pure.10Sep10.linux.1 - The energy is conserved in Column 2 and 4 of the output. But not in the other versions like
lammps-8Jan13/examples/srd/log.srd.pure.9Jan12.linux.1 and lammps-14Oct12/examples/srd/log.srd.pure.9Jan12.linux.1. The temperature of the srd fluid is supposed to be constant at 1.0 but it has decreased from 1.0 to 0.75 within 5000 steps.

Could you please look in to it.

Thanks.

2

I'll check ...

Steve

3

The reason for the different behavior is that the current version
is more careful about rescaling SRD velocities that become too
large, due to the velocity rotation operation. This protects against
the chance of losing particles, but does bleed kinetic energy out
of the system slowly. You can either thermostat via the tstat
keyword, or you can turn off the rescaling via a rescale keyword
I just added in an 18Jan patch. E.g. setting rescale no or rescale collide
will disable rescaling during rotations in a pure SRD fluid, and
then the temperature will not drop.

Steve