Hi guys,
Recently I need to recheck some of my old computations and found surprisingly that
a continuing run with the binary restart file and the data file converted from the
same binary file yields large difference in the total potential energy and the pressure
tensors, no matter what version of lammps is used. The restart run from the binary file
gives nearly identical result comparing to the last step info from the previous run.
Since I need to run the job in a different machine while I don’t have the old source code
of lammps now, I cannot run from the binary restart file on it. A conversion from the binary
file to a data file is necessary yet the large difference is totally not expected.
I checked the converted data file, some parts of the source code of lammps, but cannot
figure out the source of these differences; the PI problem with the restart2data I raised
yesterday does not matter here.
Attached please find an input script, a converted data file, as well as the “Info.dat” which
compares the thermo info from different runs.
Any clue is appreciated.
Thanks.
Lingti
Info.dat (1.02 KB)
data.pos (2.02 MB)
in.alkene (1.36 KB)