ERROR on proc 0: Out of range atoms - cannot compute PPPM (../pppm.cpp:1937)

LAMMPS LAMMPS Mailing List Mirror
1

Dear Giacomo,

Thanks a lot for your reply. I checked the forum threads that mentioned similar errors and adjusted my input files so that

  1. Pdamp parameter is sufficiently big (1000 for 1fs timesteps)
  2. neighbouring lists are updated often ( neigh_modify delay 0 every 1 check yes)

I have check the dump files and pressure seems to fluctuate very wildly (even in the NPT simulation) between positive and negative values.

As far as clumping goes, we expect it because glycerol monooleate molecules form reverse micelles.

I will run a new simulation and make the box smaller. I will also run a simulation with heptane only.

Do you have any more suggestions?

Thank you,

Joanna

2

This error is almost always due to bad dynamics, as

indicated by

Can you run the system stably with NVE, with a small

simple script? If you cannot, then running with NPT

is just masking the real issue.

Steve

3

Hi Steve,

Thank you for your reply.

I managed to run it stably with NVE ensemble for 20ps in one simulation, and in NVT ensemble for 30ps in the other simulation.
Then I changed the ensemble to NPT and the simulation stopped due to pppm, out of range atoms errror.

I realise it’s hard to say in general, but do you think I gave my system enough time to equilibrate before running NPT?

I am trying to simulate bulk liquid (mixture of 360 heptane molecules and 40 glycerol monooleate molecules) under standard conditions.

What order of ensembles do you suggest for my purposes?

Are there any parameters in my input files that are worth looking into?

Thank you very much for your help in advance,

Joanna

4

Hi Steve,

Thank you for your reply.

I managed to run it stably with NVE ensemble for 20ps in one simulation,
and in NVT ensemble for 30ps in the other simulation.
Then I changed the ensemble to NPT and the simulation stopped due to pppm,
out of range atoms errror.

I realise it's hard to say in general, but do you think I gave my system
enough time to equilibrate before running NPT?

​a few 10s of picoseconds it not much, particularly when your initial
system is far away from equilibrium.
sometimes it can also help to start with a minimization, if your initial
geometry has high potential energy.​

it would be helpful to see some of your logs, e.g. the evolution of various
energies and particularly the pressure.
the PPPM out-of-range error happens most often with systems that expand or
shrink a lot. there are multiple possible reasons for that.
​most common are: bad initial geometry, incorrect topology, bad/incorrect
force field parameters. typos etc.​

I am trying to simulate bulk liquid (mixture of 360 heptane molecules and
40 glycerol monooleate molecules) under standard conditions.

What order of ensembles do you suggest for my purposes?

​it is usually best to determine good parameters with a simpler system,
i.e. two pure compounds, and then try the mixture later.​

Are there any parameters in my input files that are worth looking into?

​*everything*

axel.​