Hello,
I’ve been using lammps (version 17 Dec 2013) to run 50 ns simulations with a 10 fs timestep and have been generating .xtc files in order to track the trajectory. The unit system being used is real (which uses fs). For the first ~33.5 ns, the time progresses correctly at 10 fs increments of time; however, beyond this point, the time begins to increase in increments of 8 and 12 fs. I’ve run simulations of several different systems and have consistently run into the same problem. Is this simply a bug in earlier versions of lammps, or is it indicative of a larger problem within my system?
Thank you,
beyond this point, the time begins to increase in increments of 8 and 12 fs.
how do you know this? Are you doing some post-processing of the XTC file
which extracts the timestep counter?
Steve
Hello,
I've been using lammps (version 17 Dec 2013) to run 50 ns simulations with a
10 fs timestep and have been generating .xtc files in order to track the
trajectory. The unit system being used is real (which uses fs). For the
first ~33.5 ns, the time progresses correctly at 10 fs increments of time;
however, beyond this point, the time begins to increase in increments of 8
and 12 fs. I've run simulations of several different systems and have
consistently run into the same problem. Is this simply a bug in earlier
versions of lammps, or is it indicative of a larger problem within my
system?
the .xtc dump style didn't use correct time units until version 21 Jan 2015.
best to disregard the time information completely and only use the
number of timesteps as indicator of time progression.
you may also want to consider upgrading to a newer version for future
simulations, not only for this one bugfix but for many other
improvements.
axel.