LAMMPS developers,
I wrote a fix that integrates the equations of motion using the Gaussian isokinetic thermostat (equations of motion shown in doi.org/10.1063/1.473273 and doi.org/10.1063/1.1534582). The code is attached, and I’m wondering if there would be any interest in my pushing this to the repo.
If so, I’d need to fix up a few things before it could actually be used in the LAMMPS distribution. I haven’t implemented the code to work with the RESPA integrator, and I’m also assuming that the group it works on is “all” (not sure if this integrator works on a system subset). For my purposes, I don’t care about these things, but if there’s community interest in having this option available in LAMMPS, I could spend some time (not right now, but in a few weeks) working on them.
fix_nvk.cpp (6.95 KB)
fix_nvk.h (1.44 KB)
LAMMPS developers,
I wrote a fix that integrates the equations of motion using the Gaussian
isokinetic thermostat (equations of motion shown in doi.org/10.1063/1.473273
and doi.org/10.1063/1.1534582). The code is attached, and I'm wondering if
there would be any interest in my pushing this to the repo.
If so, I'd need to fix up a few things before it could actually be used in
the LAMMPS distribution. I haven't implemented the code to work with the
RESPA integrator, and I'm also assuming that the group it works on is "all"
(not sure if this integrator works on a system subset). For my purposes, I
don't care about these things, but if there's community interest in having
this option available in LAMMPS, I could spend some time (not right now, but
in a few weeks) working on them.
you can also add checks, that disallow use with groups other than
"all" or r-RESPA integrators. there are plenty precedents for that.
axel.