Input script questions

LAMMPS LAMMPS Mailing List Mirror
1

Hello,

I am writing a LAMMPS input script for a dense polymeric system. So I am using the nve\limit command to limit the maximum amount my atoms can travel in a timestep, so that they do not fly across the simulation box. Over time, I want to increase the maximum distance they can travel in a time step (increment the nve\limit fix) so that my system can reach equilibrium. Do you think that the following script will work? I have tried some diagnostics on it, and it seems to work, but I would like to make sure. I unfix the old nve limit and put in a new one, and then I use the run upto ___ command to tell it how many timesteps to run with the new nve limit fix, until I delete it and put in a new one that increments the maximum amount the atoms are allowed to move, and then put in another run upto command, all until my simulation is over. Every million timesteps I unfix the old nve/limit and put in a new one, and run upto another million timesteps. Do you think that I should use the run start/stop command instead of the run upto command, or will this work fine for my purposes?

Thank you,

Here is the script:

units lj

atom_style molecular

read_data data12424.1.000000

pair_style lj/cut 2.5

pair_coeff * * 1.000000 1.000000

bond_style fene

special_bonds fene dihedral no

bond_coeff * 30.000000 1.500000 1.000000 1.000000

fix 1 all nve/limit 0.100000

fix 2 all langevin 1.000000 1.000000 1 48279

timestep 0.005000

dump MyDump all custom 1000 dump12424.1.000000.txt id x y z

run 1000000 upto

unfix 1

fix 1 all nve/limit 0.200000

run 2000000 upto

unfix 1

fix 1 all nve/limit 0.300000

run 3000000 upto

unfix 1

fix 1 all nve/limit 0.400000

run 4000000 upto

unfix 1

fix 1 all nve/limit 0.500000

run 5000000 upto

unfix 1

fix 1 all nve/limit 0.600000

run 6000000 upto

unfix 1

fix 1 all nve/limit 0.700000

run 7000000 upto

unfix 1

fix 1 all nve/limit 0.800000

run 8000000 upto

unfix 1

fix 1 all nve/limit 0.900000

run 9000000 upto

unfix 1

fix 1 all nve/limit 1.000000

run 10000000 upto

unfix 1

fix 1 all nve

run upto 1000000000

2

Hello,

hi

I am writing a LAMMPS input script for a dense polymeric system. So I am
using the nve\limit command to limit the maximum amount my atoms can travel
in a timestep, so that they do not fly across the simulation box. Over

if people could predict the outcome of a simulation by just looking at
the input file, we would not need a simulation program? :wink:

time, I want to increase the maximum distance they can travel in a time step
(increment the nve\limit fix) so that my system can reach equilibrium. Do

i don't think that this is a viable idea. if your atoms are too close, then you
can just start with a minimization and then you should be fine. if you atoms
do overlap in a way that bonds are crossing, then i would suggest to use the
"soft" potential instead. you can gradually increase the potential depth over
time using fix adapt.

you think that the following script will work? I have tried some
diagnostics on it, and it seems to work, but I would like to make sure. I
unfix the old nve limit and put in a new one, and then I use the run upto
___ command to tell it how many timesteps to run with the new nve limit fix,
until I delete it and put in a new one that increments the maximum amount
the atoms are allowed to move, and then put in another run upto command, all
until my simulation is over. Every million timesteps I unfix the old
nve/limit and put in a new one, and run upto another million timesteps. Do
you think that I should use the run start/stop command instead of the run
upto command, or will this work fine for my purposes?

i don't see the point of your question. if you unifx a fix, it is
gone, literally.
the class that represents it will be deleted and thus its storage and
state freed.
if you create a new fix, even if it is the same kind of fix, it will be freshly
initialized. so what exactly would be the problem?

cheers,
    axel.