Equilibrium problem

Dear Sir,
Good noon. I am trying to do equilibrium of a reactive force field system. In my system Hyaluronan is emerged in water. At first I have fixed Hyaluronan while water goes to equilibrium process for 2.5ps, then I have equilibrated total system for 30ps freeing Hyaluronan. However, after plotting it is showing total energy is still fluctuating in fact increasing. This is my equilibrium command

fix freeze HA setforce 0.0 0.0 0.0

fix mynpt all npt temp 10.0 300.0 100.0 iso 1.0 1.0 1000.0 #drag .01

fix 2 all qeq/reax 10 0.0 10.0 1.0e-6 reax/c

run 10000
write_data haslab_1.data

unfix freeze

run 120000
write_data haslab_2.data

should I run the equilibrium for much higher time step? Thank you.

Best Regards,
Khandakar Abu Hasan Al Mahmud

Dear Sir,
Good noon. I am trying to do equilibrium of a reactive force field system.
In my system Hyaluronan is emerged in water. At first I have fixed
Hyaluronan while water goes to equilibrium process for 2.5ps, then I have
equilibrated total system for 30ps freeing Hyaluronan. However, after
plotting it is showing total energy is still fluctuating in fact increasing.
This is my equilibrium command

fix freeze HA setforce 0.0 0.0 0.0

fix mynpt all npt temp 10.0 300.0 100.0 iso 1.0 1.0 1000.0
#drag .01

fix 2 all qeq/reax 10 0.0 10.0 1.0e-6 reax/c
run 10000
write_data haslab_1.data
unfix freeze

run 120000
write_data haslab_2.data

should I run the equilibrium for much higher time step? Thank you.

this is incomplete input and thus mostly useless as lots of crucial
information is missing. also some fluctuations are normal.
also the input fragments you quote are actually not showing an
equilibration simulation but a thermalization with linear increase in
temperature.

a few suggestions anyway:
- don't use fix npt unless you really have to. if you can adjust your
system to have the desired density, do equilibration with constant
volume.
- for thermalization and initial equlibration, using a dissipative
thermostat (langevin, temp/csld, gle, gld, combined with fix nve, as
needed) is more effective, as they accelerate the equipartitioning of
kinetic energy.
- you should check for energy conservation *without* a thermostat. the
most common reason for bad energy conservation is too large a time
step
- monitor your total system momentum (and reset whenever you restart
during thermalization and equilibration).
- carefully check whether your potential file is suitable for the
system and the conditions you are simulating. reactive force field
parameterizations are usually less transferable than simpler
(non-reactive) force field parameter sets.

axel.