[lammps-users] bad fluctuation in temperature and pressure.

Dear all,

I am trying to simulate a system that contain totally 53 atoms.
I want to do this in NPT integrator to fix pressure and temperature, but there is bad fluctuation in temperature and pressure. They do not reach equilibrium at all.


units metal
atom_style full
lattice fcc 1.5
region AA block -10 10 -10 10 -10 10
bond_style harmonic
angle_style charmm

pair_style hybrid tersoff lj/charmm/coul/long 8 10 lj/cut 8.5

read_data 1water+cnt.data

pair_coeff * * tersoff SiC.tersoff NULL NULL C
pair_coeff 1 1 lj/charmm/coul/long 0.002012368 0.4000 0.002012368 0.4000 # H-H
pair_coeff 1 2 lj/charmm/coul/long 0.003625732 1.7753 0.003625732 1.7753 # H-O
pair_coeff 2 2 lj/charmm/coul/long 0.006596577 3.11507 0.006596577 3.11507 # O-O
pair_coeff 2 3 lj/cut 0.004475784 3.33256 #by mixing rule O-C
pair_coeff 1 3 lj/cut 0.002471656 1.97503 #by mixing rule H-C

kspace_style pppm 1e-4
special_bonds charmm

group A id < 4
displace_atoms A move 3 0 0

velocity all create 300.0 4928459 dist gaussian

fix 1 all npt 300.0 300.0 .1 xyz 1.013 1.013 .1

timestep 2e-4

Step Temp E_pair E_mol TotEng Press Volume
0 300 -290.57346 82.121814 -206.43518 9198.6078 5780.5987
10000 105.83701 -295.56909 1.3271611 -293.53054 212.23416 6411.949
20000 59.232151 -295.70237 0.50530395 -294.79893 -5.3970655 6408.4155
30000 1263.1387 -295.77179 31.049159 -256.23241 937.28849 6402.123
40000 994.25011 -295.80329 24.694039 -264.42637 910.39998 6349.1399
50000 826.29982 -295.82804 11.278439 -278.99561 -1941.0155 6327.8856
60000 50.904298 -295.83968 1.717065 -293.78046 350.40605 6296.2202
70000 44.448121 -295.84961 9.300208 -286.25065 -1352.389 6258.3532
80000 138.23974 -295.85634 2.2126485 -292.71451 25.22238 6217.2142
90000 100.94822 -295.85811 1.6218142 -293.55777 60.442025 6178.9523
100000 187.80415 -295.85793 3.6825518 -290.91305 1325.2804 6115.6812
110000 190.04818 -295.86386 3.942642 -290.64381 919.11409 6085.4201
120000 115.04161 -295.86295 2.3866795 -292.70302 114.53183 6028.547
130000 44.53459 -295.86817 10.471168 -285.09766 -2692.2584 6001.0755
140000 23.976559 -295.87068 0.52905999 -295.18046 202.49912 5950.9218
150000 369.94394 -295.87222 4.1558524 -289.22978 840.59859 5896.9707
160000 114.05657 -295.8751 2.6478946 -292.46057 480.45811 5856.8448
170000 43.283379 -295.87199 0.8125509 -294.7685 397.16651 5799.379
180000 41.162328 -295.87483 0.82393793 -294.77422 122.92849 5766.6466

Regards,
Yasti

Dear all,

I am trying to simulate a system that contain totally 53 atoms.

how can you even expect that pressure and temperature will
not wildly fluctuate with so few atoms??

there are many discussions and suggestions on how to do
npt so that fluctuations are minimal in the mailing list archive.
please help yourself to a look. thanks,
     axel.

I want to do this in NPT integrator to fix pressure and temperature, but
there is bad fluctuation in temperature and pressure. They do not reach
equilibrium at all.

[...]

Dear Yasti,
As Axel mentioned and have said repeated before as well, please try to proceed step by step, constantly figuring out what are the reasoning behind what your are getting and if that makes logical sense. Please don’t expect LAMMPS to perform miracles in one go.

Few questions on the contrary to you

  1. Why did you start with such a small system?
  2. Why did you choose to NPT right from the start?
  3. Why did you expect that that temperature and pressure will not fluctuate much, at the first place?
  4. Why did you not minimize the system in the beginning to make sure that initial forces are not huge? I can see those right from the temperature evaluation?
  5. It is also a little odd to see cutoff at 8 Angs.

Best Regards,
Vikas