NPT fix_nh questions

LAMMPS LAMMPS Mailing List Mirror
1

Hello all,

I’m trying to understand how LAMMPS integrates with respect to NPT ensemble (not NPH or NVT); and I’m not doing RESPA. I have also read the fix_nh documents and its reference but it still seems unclear to me. I have a few questions, particularly about fix_nh.cpp, which is used to address NPT ensemble.

  1. Is it true that in fix_nh.cpp, pstat_flag = tstat_flag = 1 for NPT ensemble?

  2. Where is exactly the set of governing equations for NPT ensemble in LAMMPS? Both Shinoda’s paper and Tuckerman’s paper include this set, but they are not exactly the same. Are they equivalent in some ways that I can’t see?

  3. What is mpchain in fix_nh.cpp?

Thank you for your time and help.

Regards,
Anh Tran

PhD student

Multiscale Systems Engineering Research Group

Georgia Institute of Technology

2

Hello all,

I'm trying to understand how LAMMPS integrates with respect to NPT ensemble
(not NPH or NVT); and I'm not doing RESPA. I have also read the fix_nh
documents and its reference but it still seems unclear to me. I have a few
questions, particularly about fix_nh.cpp, which is used to address NPT
ensemble.

1) Is it true that in fix_nh.cpp, pstat_flag = tstat_flag = 1 for NPT
ensemble?

yes. please see the constructor in fix_npt.cpp.

2) Where is exactly the set of governing equations for NPT ensemble in
LAMMPS? Both Shinoda's paper and Tuckerman's paper include this set, but
they are not exactly the same. Are they equivalent in some ways that I can't
see?

that is a question for aidan.

3) What is mpchain in fix_nh.cpp?

the length of the nose-hoover chain for the barostat, same as mtchain
for the thermostat.
the meaning of most class members variables can be seen from the
comments in the corresponding header files. check out fix_nh.h

axel.

3

Hello all,

I'm trying to understand how LAMMPS integrates with respect to NPT ensemble
(not NPH or NVT); and I'm not doing RESPA. I have also read the fix_nh
documents and its reference but it still seems unclear to me. I have a few
questions, particularly about fix_nh.cpp, which is used to address NPT
ensemble.

1) Is it true that in fix_nh.cpp, pstat_flag = tstat_flag = 1 for NPT
ensemble?

yes. please see the constructor in fix_npt.cpp.

2) Where is exactly the set of governing equations for NPT ensemble in
LAMMPS? Both Shinoda's paper and Tuckerman's paper include this set, but
they are not exactly the same. Are they equivalent in some ways that I can't
see?

that is a question for aidan.

The LAMMPS time integration scheme differs slightly from that of
Tuckerman in how barostat coupling is combined with rRESPA. The
LAMMPS scheme is fully defined by the symmetric Trotter factorization
given on the fix_nvt doc page. This factorization can be converted to
pseudocode by the method of direct translation (Martyna, Tuckerman ,
Tobias, and Klein, 1996).