The temperature keeps rising with NVE, pair_style dpd, to simulate an amphiphilic polymeric system.

LAMMPS LAMMPS Mailing List Mirror
1

thank you for your kind responses. I’ve got some comments on my problem.

  1. As the manual indicates, dpd/tstat invokes pairwise interactions which are equivalent to the non-conservative portion of the DPD force field. I think this is not suitable for my model.
  2. I found the fix temp/rescale command can rescale the system temperature well, but I am not sure if this command will cause inappropriate DPD interactions for my model?
  3. I found from literatures that many people use NVT ensemble in their simulations. However, as "pair_style dpd +fix nvt " can not be used because of the repeated thermostats and “pair_style dpd/tstat + fix nvt” do not consider Fc, I do not know which command can be used to conduct “dpd + NVT” simulation. Hope you can give me some suggestions based on your experience. Thanks!

zhikun

2

thank you for your kind responses. I've got some comments on my problem.
1) As the manual indicates, dpd/tstat invokes pairwise interactions which
are equivalent to the non-conservative portion of the DPD force field. I
think this is not suitable for my model.
2) I found the fix temp/rescale command can rescale the system temperature
well, but I am not sure if this command will cause inappropriate DPD
interactions for my model?

no. fix temp/rescale will only mask the real problem. your simulation
has to conserve energy with pair style dpd and fix nve. if it doesn't,
something is wrong.

3) I found from literatures that many people use NVT ensemble in their
simulations. However, as "pair_style dpd +fix nvt " can not be used because
of the repeated thermostats and "pair_style dpd/tstat + fix nvt" do not
consider Fc, I do not know which command can be used to conduct "dpd + NVT"
simulation. Hope you can give me some suggestions based on your experience.

dpd includes the thermostat, so fix nve is adequate. technically, i
don't think that using the term NVT ensemble is a correct for DPD in
general, but that is a different issue. as steve pointed out, there
are more complex DPD models in the USER-DPD package, but your problem
seems to be a very fundamental one. i believe that you will first need
to debug and identify what is causing the energy issue. but without a
data file, that is impossible to tell.

axel.