boundary 'p p p' and 's p p' , giving same potential energy

Respected Sir,

I carried out two run, one with p p p and other with s p p boundary conditions.
The values of potential energy is same in both cases. How this is possible?

The relevant files are attached.

input2 is for ‘p p p’

input4 is for ‘s p p’

Please reason this out. Thankful in advance.

input2 (2.34 KB)

out2.dat (8.46 KB)

input4 (2.34 KB)

out4.dat (9.78 KB)

Respected Sir,

I carried out two run, one with p p p and other with s p p boundary
conditions.
The values of potential energy is same in both cases. How this is possible?

please study your outputs *very* carefully. the answer to your
question is in there.

axel.

I think, that the ‘s p p’ run is using the boundary settings of restart file i.e. ’ p p p’. Is it the reason?

I think, that the 's p p' run is using the boundary settings of restart file
i.e. ' p p p'. Is it the reason?

you are a trained researcher, are you not? then why do you need to ask
about something that should be well within your means to answer
properly yourself?

you have the evidence before you, and you can confirm it by studying
the documentation and making tests to verify what the documentation
states.
there also is the "info" command LAMMPS that allows introspection of
the state of the simulated system and active LAMMPS settings.

axel.

Sir,

I found that Neither NVT,NPT nor box/relax works for ‘non periodic’ system. But since it was working for ‘s p p’ , so I figured out that it must be using restart file boundary conditions.

I just mailed for confirmation of my finding.

Sir,
I found that Neither NVT,NPT nor box/relax works for 'non periodic' system.

that is an *incorrect* statement.
you can you use fix nvt with *any* boundary condition and also both
fix npt and fix box/relax can be used with (partially) non-periodic
boundaries for as long as they are only coupled to periodic
directions.

But since it was working for 's p p' , so I figured out that it must be
using restart file boundary conditions.

it is sad to read, that you come to the correct conclusion based on
incorrect assumptions and a disappointing neglect of warnings in your
output as well as ignorance of the LAMMPS manual (the documentation
for the read_restart command lists in detail which settings are
imported from reading a restart).

if you look in your output, LAMMPS is *telling* you that it will
override the boundary settings that you have with those from the
restart file with a BIG FAT WARNING message.
you can also confirm this properly through studying the documentation
more carefully than you apparently did and issuing the info command
with suitable settings at suitable places in your input.

I just mailed for confirmation of my finding.

...and i just mailed to confirm, that your conclusions - while leading
to the correct answer - are incorrect since they are based on
incorrect assumptions. hopefully this will help to remind people of
the *correct* way to assess such situations, how important it is to
take warning messages seriously, and that wild guesses are rarely a
good way to solve problems, even if they sometimes leads to a correct
answer.

axel.