questions about the LAMMPS peptide example...

I have been experimenting with the peptide example since I am interested in modeling water inside a polymer and this example seems to be
the closest one. The example uses NVT and the temperature cycles around the 275 K as expected with the thermostat. Upon lowering the temperature to 75 K the simulation honors this temperature as well. Removing the " tchain 1" or changing the number from 1 to 10 seems to have no effect even for a run lasting 30000 time steps. The peptide example uses real units so pressure is in atmospheres. The pressure varies wildly during the simulation moving from many thousands of atmospheres positive to negative. The number of atoms (N), volume (V) and temperature (T) are held constant while pressure is allowed to vary.

Given the modeling assumptions of CHARMM and harmonic force fields and the pair_style for this example is it possible to run it using NPT and get realistic pressures in the 1 ATM range at 275 K? Do I need to start at very low temperature and pressure and run for a very long time to get
convergence to a stable T and P?

Are there examples available for polymer systems at room temperature and pressure?

I tried the following for a NPT run:

fix 1 all npt temp 275.0 275.0 100.0 iso 1.0 1.0 1000.0 tchain 1 pchain 1

I had to specify the following

fix 2 all shake 0.0001 10 100 b 4 6 8 10 12 14 18 a 31

prior to the fix statement above.

The pressure varies wildly in this case as well.

Thank you,

Bruce

in.peptide (630 Bytes)

data.peptide (292 KB)

bruce,

I have been experimenting with the peptide example since I am interested in
modeling water inside a polymer and this example seems to be
the closest one. The example uses NVT and the temperature cycles around the
275 K as expected with the thermostat. Upon lowering the temperature to 75 K
the simulation honors this temperature as well. Removing the " tchain 1" or
changing the number from 1 to 10 seems to have no effect even for a run
lasting 30000 time steps. The peptide example uses real units so pressure is

changing the nose-hoover thermostat chain length _does_ have an effect,
but one that is not immediately visible.

in atmospheres. The pressure varies wildly during the simulation moving from
many thousands of atmospheres positive to negative. The number of atoms
(N), volume (V) and temperature (T) are held constant while pressure is
allowed to vary.

that is not entirely correct. the instantaneous temperature does fluctuate.

Given the modeling assumptions of CHARMM and harmonic force fields and the
pair_style for this example is it possible to run it using NPT and get
realistic pressures in the 1 ATM range at 275 K? Do I need to start at very
low temperature and pressure and run for a very long time to get
convergence to a stable T and P?

instantaneous temperature and pressure _do_ fluctuate.
only their averages are supposed to be constant.
pressure fluctuates more than temperature in a liquid
due to the low compressibility. this is a statistical
effect, i.e. if you simulate a larger system, then the
fluctuations are smaller. please have a look at a text
book on MD and statistical mechanics to see more
detailed explanations for these phenomena.

Are there examples available for polymer systems at room temperature and
pressure?

I tried the following for a NPT run:

fix 1 all npt temp 275.0 275.0 100.0 iso 1.0 1.0 1000.0 tchain 1 pchain 1

I had to specify the following

fix 2 all shake 0.0001 10 100 b 4 6 8 10 12 14 18 a 31

prior to the fix statement above.

The pressure varies wildly in this case as well.

that is to be expected.

cheers,
    axel.