No change to cell during energy minimization and regardless of order

LAMMPS LAMMPS Mailing List Mirror
#1

Dear LAMMPS Help List,

Good morning. I have tried several times on my own to energy minimize a hectorite sample. I have tried the energy minimization command on its own and then NPT at (about) 0 K. However, I realize that fix box/relax is likely the best bet (particularly after reading the manual and help list), but I see no change to my cell when I do that. The energy remains constant the entire time as well. After going through the old posts to the mail lists, I saw someone suggested doing a fix box/relax and an energy minimization after the fix. However, when I try this, I get the same, no change effect to my cell. Do you have any other suggestions? Below, I have included the current order of my input file. I would be grateful for any advice you could give me.

Fraction of in.lammps file

#Dynamics
timestep 0.5
fix 1 all box/relax aniso 0.0 vmax 0.001
#fix 2 all temp/berendsen 0.0001 0.0001 0.0 # Not Necessary when NVT is used
#compute 1 hect msd
#compute 2 hect rdf 1000 9 9 10 9 10 10
#fix 2 all ave/time 10 10 100 c_2 ave running start 10 file tmp.rdf mode vector
#fix 3 all ave/time 10 10 100 c_1 ave running start 10 file tmp.msd mode vector
dump 1 all xyz 1000 dump.water.xyz
dump 2 all custom 1000 dump.all id type x y z
dump 3 all atom 100 dump.atom
thermo 100
thermo_style custom etotal pe ke evdwl ecoul epair ebond eangle edihed eimp elong etail temp press enthalpy
thermo_modify line multi
#thermo_style multi custom cpu
#thermo_style emol elong etail c_1 c_2 c_3 fix_1 fix_2
run_style verlet
run 250000
restart 1000 restart_hect_min.0
write_restart restart_hect_min.0

#Minimization

min_style cg
minimize 1.0e-6 1.0e-8 500 1000
write_restart restart_water-min.300

Thank you very much, and have a great day.
Christin

#2

Dear LAMMPS Help List,

dear christin,

Good morning. I have tried several times on my own to energy minimize a
hectorite sample. I have tried the energy minimization command on its own
and then NPT at (about) 0 K. However, I realize that fix box/relax is
likely the best bet (particularly after reading the manual and help list),
but I see _no change_ to my cell when I do that. The energy remains

lets be clear about this:

running an MD with fix npt at 0K will not move any atoms, you
will have to do a simulated annealing protocol and slowly remove
kinetic energy, either by programming a reasonably shallow
temperature gradient or using fix nph in combination with fix viscous
after a proper initial thermalization.

constant the entire time as well. After going through the old posts to the
mail lists, I saw someone suggested doing a fix box/relax and an energy
minimization _after_ the fix. However, when I try this, I get the same, no
change effect to my cell. Do you have any other suggestions? Below, I have
included the current order of my input file. I would be grateful for any
advice you could give me.

it is near impossible to give advice on partial data.
when you have trouble with something not working
as expected/advertised, you almost always need to
produce a (small!) example input that reproduces
the "bad" behavior. the cause for your problem may
lay in some part of the input that you don't provide.

for example, i doubt that there is any impact of the
order of the input on your issue. it has to be something
much more fundamental, but that is very difficult to
tell from the little data you provide.

cheers,
    axel.