Moving window calculation in LAMMPS

Dear LAMMPS team,

is there a way to implement a moving window calculation, i.e. to make atoms enter simulation box from one side, move through the box and go out from the opposite boundary, where they are destroyed?

Oleg Sergeev,
postgrad student,
MIPT (Moscow)

Hi Oleg,

I think a combination of fix deposit (with velocities set) and fix evaporate (with a large N) with appropriate regions set might help. I’ve never used it, but hope this helps.

  • Oscar

Thanks for the suggestion,

I think this won't work in our case, because we'd like to simulate complex crystalline structure.

Oleg.

15.10.2012, 21:07, "Oscar Cushing" <[email protected]>:

Thanks for the suggestion,

I think this won't work in our case, because we'd like to simulate complex crystalline structure.

would you mind to explain, why you'd need to do the
"moving window" then? that might help to provide a
suitable simulation strategy.

good help depends on good information. :wink:

cheers,
    axel.

Ok :slight_smile:

It is supposed that inside the box will be a (initially) plane reaction front, and we are interested in whether it will change its shape due to instabilities. So we need a large cell in the directions normal to that of the front propagation and large simulation time, and it would be too costly to simulate non-reacting (yet) matter, too.

And the reaction takes place in a crystal.

Oleg.

16.10.12, 08:56, "Axel Kohlmeyer" <[email protected]>":

The fix append/atoms command may do what you want.

Steve

Thank you, I'll try it out. It looks appropriate.

Oleg.

Dear Dr. Plimpton,

I think there is a typo in 16th line of SHOCK/fix_append_atoms.cpp. LAMMPS could not recognize the fix until append_atoms is replaced with append/atoms in that line.

Oleg.

16.10.12, 22:24, "Steve Plimpton" <[email protected]>":

Sorry, not .cpp, but .h.

Oleg.

17.10.12, 13:52, "[email protected]" <[email protected]>":