Dear all,
As the subject says I am having troubles with keeping the temperature constant using nvt or npt ensamle. I am running a simulation on polyethylene provided in the benchmark section with AIREBO potential. Instead of the temperature being fixed at 296K it drops to around 160-170K and keeps fluctuating but never going anywhere near 296K. It would be great if somebody could help me fix this.
The input file is pasted below:
AIREBO polyethelene benchmark
units metal
atom_style atomic
read_data data.airebo
replicate 1 1 40
neighbor 0.5 bin
neigh_modify delay 5 every 1
pair_style airebo 3.0 1 1
pair_coeff * * CH.airebo C H
velocity all create 296.0 761341
fix 1 all nvt temp 296.0 296.0 100.0
fix 2 all phonon 10 5000 20000 map_z.in airebo_z nasr 100
timestep 0.0001
thermo 100
run 20000000
Dear all,
As the subject says I am having troubles with keeping the temperature
constant using nvt or npt ensamle. I am running a simulation on polyethylene
provided in the benchmark section with AIREBO potential. Instead of the
temperature being fixed at 296K it drops to around 160-170K and keeps
fluctuating but never going anywhere near 296K. It would be great if
somebody could help me fix this.
your thermostat time constant is 100ps and your total simulation time is 2ns.
please see the fix nvt docs for what constitutes a good choice of
parameters for fix nvt/npt
fix nvt/npt are not very good thermostats to heat or cool a system
quickly and reach equipartioning of the kinetic energy (as a
prerequisite for equilibrium). it is generally recommended to start
with fix nve+langevin, for systems far from equilibrium it may be
necessary to do this in stages and first use a short time constant and
then a longer and then switch to fix nvt/npt.
as for the actual instantaneous temperature, please see the many
previous discussions on finite size effects and temperature/pressure
fluctuations.
also, be careful with benchmark inputs, they may not always represent
physically meaningful systems, but designed to stress the performance
of certain components of a code.
axel.
Dear Axel,
Thank you, I fixed it. Now I would appreciate help with another thing. I try using topotools to visualize the structure that I am feeding into the simulation, but I couldn’t use it with the provided .data file,because there is no bonds information provided in it for polyethylene. How do I go about this then?