[lammps-users] LAMMPS Question


Is there a command for energy minimization in the new LAMMPS version? I noticed there was a command "minimize" in the older version but I don't find this anymore. Any help is appreciated - in particular if someone could tell me a way to carry out energy minimization with CG or similar methods.


Hello Markus,

in the Fortran version there is still
the minimize command and here is a sample
of the commands that you have to include
in the input file

min flag 1
minimize 0.001 1000 50000

I am quite sure that the c++ version does not
have it. As for conjugate gradient it is one
of the coming attractions for Lammps. So for now
either you have to use Numerical Recipes routines
to code it in or use another code, one of them
being gromacs.

Hope these helps,


Hi Markus,

Currently there isn't any minimization capability within the C++ version
of LAMMPS, but I think they are planning to add conjugate gradient
sometime soon. There was a recent email from Paul Crozier on the list (I
think last Thursday, but I can't seem to find the link on the sourceforge
archive page).

What I do is a little ad hoc but seems to work for me, at least to get a
decent configuration to equilibrate (I'm using a coarse-grain model as

I just set the timestep really low (1fs) and do a whole bunch of 10-50
step NVE runs, setting the velocities of all the atoms to zero between
each run.