local stress calculation after subtracting bias

Hi everybody,

From the previous discussions in this forum regarding the shock simulation, I have learned a lot. Thanks to those folks for their valuable remarks. Anyway, using “momentum mirror” method by the “wall/reflect” & “velocity set sum yes” command, I have prepared the input script where everything is fine because I was able to reproduce the earlier results. Temperature was computed using “temp/com” (thanks to Shan and Aidan for their suggestions in this forum) which eventually excludes COM biasing but in case of pressure, how can I compute that specially with spatial averaging ?
Is there any way within Lammps environment such as like “temp/com” ?

Is there anybody who have computed local stress after subtracting COM bias ?
Any help would be greatly appreciated.

Thanks,

Liu

The issue of excluding COM velocity contribution to per-atom stress is
sufficiently complicated that LAMMPS does not support it directly. It
is best handled in post-processing. However, it is possible to exclude
the kinetic contribution entirely using the virial keyword in compute
stress/atom. The resultant per-atom virial stress can then be summed,
time-averaged, etc., and it can also be combined with a kinetic
contribution obtained from temp/com to estimate the correct local
stress.

Aidan

Hi,

Aldan wrote: It
is best handled in post-processing. However, it is possible to exclude
the kinetic contribution entirely using the virial keyword in compute
stress/atom. The resultant per-atom virial stress can then be summed,
time-averaged, etc., and it can also be combined with a kinetic
contribution obtained from temp/com to estimate the correct local
stress.

Aldan valuable comment gave me great ideas =) .

Just wanted to say Thanks …

Oscar G.