Question: nvt/shake and rigid/nvt

LAMMPS LAMMPS Mailing List Mirror
1

Dear lammps members,

2

SHAKE in LAMMPS is limited to small clusters, like water,
or CH bonds in a polymer chain.

If you can use fix shake to rigidify your model (e.g. water),

then you should use it. If you can’t b/c the molecule or cluster
is too big, then use fix rigid or fix rigid/small.

Steve

3

Thank you for the suggestion.

I did not use " fix rigid" or " fix rigid/small" since they ate NVE time stepping, and I needed

NVT equilibriation.

and I wanted to find if one is preferable over the other.

4

Thank you for the suggestion.

I did not use " fix rigid" or " fix rigid/small" since they ate NVE time stepping, and I needed
NVT equilibriation.

So I tested " fix nvt + fix shake" and “fix rigid/nvt” – both seems to work,
and I wanted to find if one is preferable over the other.

Fix shake usually allows a larger timestep.

5

Thank you.

regards

shankar

6

Hi Shankar,

fix rigid/small/nvt and other variants are underway. For now, you can use fix rigid/small coupled with fix langevin as an alternative to sample the NVT ensemble.

You can try yourself to see if the time step for rigid/small + langevin should be smaller than fix shake + fix nvt for your particular simulations.

Best,

-Trung

7

Dear Trung,

Thank you for the suggestions.

For now I am testing methods for nvt equilibriation of nitrogen molecules; later, study it for

computing drag on cnt in nitrogen flow.

I briefly looked up references on “time-integrators” ( lammps doc: Kamberaj 2005 for rigid/nvt) and others; but I am not yet well versed them and I wanted to learn from the experience of lammps users for their specific problems/issues

I will study “rigid/small + langevin” as you suggest.