(no subject)

Hi all,
I’m a new LAMMPS-user. Could you kindly explain me a question

  1. For my understanding, if I compress (or stretch) a non-periodic simulation box, its volume will be changed such that I cannot use fix nve or fix nvt. Maybe, it doesn’t make perfect sense to compress non-periodic simulation box but I am going to consider initial crack in model.
    However, when I use fix npt, an errors appear like “Cannot use fix nvt/npt/nph on a non-periodic dimension (fix_nh.cpp:315)”
    What would you advise me to do? Any suggestion will be appreciated.
    Thanks
    Nam

Hi all,
I'm a new LAMMPS-user. Could you kindly explain me a question
1. For my understanding, if I compress (or stretch) a non-periodic
simulation box, its volume will be changed such that I cannot use fix nve or

you certainly can use fix nve or fix nvt under these conditions.

fix nvt. Maybe, it doesn't make perfect sense to compress non-periodic
simulation box but I am going to consider initial crack in model.
However, when I use fix npt, an errors appear like "Cannot use fix
nvt/npt/nph on a non-periodic dimension (fix_nh.cpp:315)"
What would you advise me to do? Any suggestion will be appreciated.

the logic works the other way around. fix nve for example doesn't
mean that the box (or total energy) is forced to be unchanged
but rather that the fix doesn't make any attempt to adjust them.

what is the best way to proceed depends on what exactly
you want to do and how.

for an example of a (2d) crack simulation, you can have
a look at the input in examples/crack/

cheers,
    axel.

Special thanks for your speedy question

Cheers,

Nam

2011/12/19 Axel Kohlmeyer <[email protected]>

The error message is that you can't use a barostat
on a non-periodic dimension. You can use
a thermostat. A barostat has nothing to adjust
(i.e. the volume) if the system is non-periodic.

Steve

Thanks so much for your reply
However, if I use thermostat fix npt such as

fix 1 all npt temp 500.0 500.0 10.0

ERROR: Pressure control must be used with fix npt (fix_npt.cpp:29)

Could you please tell me how to overcome this mistake.

Otherwise, using fix nvt for non-periodic, I affraid the ensemble is not a canonical anymore. Please correct if I am wrong

Thanks in advance
Nam

2011/12/20 Steve Plimpton <sjplimp@…43…4…>

Thanks so much for your reply
However, if I use thermostat fix npt such as

fix 1 all npt temp 500.0 500.0 10.0

ERROR: Pressure control must be used with fix npt (fix_npt.cpp:29)

Could you please tell me how to overcome this mistake.

by reading the documentation. if you request to adjust
the box size according to the pressure, but don't give
a target pressure, what should the program use as a
reference. this is like turning on a thermostat without
telling it to what temperature you want it to adjust.

Otherwise, using fix nvt for non-periodic, I affraid the ensemble is not a
canonical anymore. Please correct if I am wrong

you *are* wrong. if you don't have periodic boundaries,
you won't have a canonical ensemble in any case.

it looks to me, that you are not understanding the
issue on a fundamental level. perhaps it may be
better to step back for a bit, read some stat mech
texts and MD principles books and then think about
what people like steve or me told you and then
try again. seems much more promising than making
(wrong) assumptions, guessing flags and not
understanding why lammps is barking at you.

...and you should be thankful that lammps doesn't
accept your input. in other cases, it would run,
but give you (more or less) wrong results, which
would be even worse.

axel.