Dear all,
I am working on simulation of water in CNT using LAMMPS software. To do this, I used moltemplate to build molecules of water with SPC model in zigzag CNT. I used langevin thermostat (at 298K and large damping parameter) with NVE for minimization of energy (This thermostat only applied on water molecules and not on CNT) and then I used NVT (at 298K) to run the simulation. However, the temperature of system during minimization and running the simulation reaches to 100 K and also get high pressure values which I can’t explain.
To figure out the problem, I only ran the water molecules without CNT which gave me reasonable pressure and temperature values (300K) in my simulation. I am wondering how can I approach the problem for water in CNT?
BTW, I also tried minimize command for minimization with same parameter in manual and setforce 0 for CNT, and NVT (at 298K) for running the simulation, but still I can’t get the acceptable results after running the simulation. I attached the summery of my results and main part of input file to this email.
Thanks and regards,
Hamed.
-------------------------------------------------------------input file---------------------------------------------------------------------
PREREQUISITES:
Results.pdf (339 KB)
Dear all,
I am working on simulation of water in CNT using LAMMPS software. To do
this, I used moltemplate to build molecules of water with SPC model in
zigzag CNT. I used langevin thermostat (at 298K and large damping parameter)
with NVE for minimization of energy (This thermostat only applied on water
langevin+nve with _large_ damping parameter as replacement for
minimization?? that makes no sense at all! where did you learn that?
if you want to do simulated annealing for minimization, you would use
fix nve in combination with fix viscous (and a suitably chosen
friction).
if you want to use a thermostat with high dissipation, you would use a
_small_ damping parameter. the damping parameter is a time constant
signifying the time to reaching the target temperature from an initial
excess or lack of kinetic energy, thus the larger the time constant,
the weaker the thermal coupling.
molecules and not on CNT) and then I used NVT (at 298K) to run the
simulation. However, the temperature of system during minimization and
running the simulation reaches to 100 K and also get high pressure values
which I can't explain.
To figure out the problem, I only ran the water molecules without CNT which
gave me reasonable pressure and temperature values (300K) in my simulation.
I am wondering how can I approach the problem for water in CNT?
BTW, I also tried minimize command for minimization with same parameter in
manual and setforce 0 for CNT, and NVT (at 298K) for running the simulation,
but still I can't get the acceptable results after running the simulation. I
attached the summery of my results and main part of input file to this
email.
two comments/questions:
- if you are using fix setforce, you do remove force *after*
contributions to the virial tensor (i.e. pressure) have been
accumulated. so if your CNT geometry is not minimized, it can
contribute quite a bit to the pressure. best to exclude computing
intramolecular force contributions altogether with neigh_modify for
those atoms.
- if you have a system with part of the atoms a 0K and part of the
atoms thermalized at 300K, what is the total temperature going to be?
axel.