[lammps-users] With nve still with npt moving.

LAMMPS LAMMPS Mailing List Mirror
1

Hi ,
i have a rigid construction that i want to keep it still. i used setforce with zeroes. I am using a nve integrator to the rest of the molecules. after some equilibration i change to npt but only for the moving particles. but when i am trying to visualize the system i see that the rigid construction is moving for a while and then it makes some vibrations. Why is this happening? Do i need to add something else?

fix 1 tubes rigid single
fix 2 tubes setforce 0.0 0.0 0.0
fix 3 moving nve
fix 4 moving temp/rescale 100 300.0 300.0 10.0 1.0

2

Hi ,
i have a rigid construction that i want to keep it still. i used

setforce with zeroes. I am using a nve integrator to the rest of the
molecules. after some equilibration i change to npt but only for the

if you don't want it to move, just don't include it in the
time integration.

moving particles. but when i am trying to visualize the system i see
that the rigid construction is moving for a while and then it makes
some vibrations. Why is this happening? Do i need to add something
else?

you need to read the documentation of fix npt!

particularly the part about the dilate keyword
which explains exactly how to handle the scenario
that you are describing.

fix 1 tubes rigid single
fix 2 tubes setforce 0.0 0.0 0.0

i also suggest to read the documentation for fix rigid
again. rigid means "move together" and not "don't move".
for the latter just don't time integrate those atoms.

a.

3

In order to have my rigid structure still i used setforce. i haven’t integrated it. What i was thinking is that when i use the npt the volume changes so is there any possibility that when the volume changes ,the axes change too so thats why i print different coordinates at the beginning. Is this possible?

4

2010/6/17 Jim Jim <[email protected]...>:

In order to have my rigid structure still i used setforce. i haven't
integrated it. What i was thinking is that when i use the npt the volume

fix rigid will integrate it. RTFM!

changes so is there any possibility that when the volume changes ,the axes
change too so thats why i print different coordinates at the beginning. Is
this possible?

i don't understand what you are trying to say here.

a.

5

Ok. What i can’t find at the manual is, if with the npt there is a change at the origin of the coordinat esystem and that’s why i see this difference at the values of the coordinates.

Thanks again
Jim

6

2010/6/17 Jim Jim <[email protected]...>:

Ok. What i can't find at the manual is, if with the npt there is a change at
the origin of the coordinat esystem and that's why i see this difference at
the values of the coordinates.

origin of the coordinate system has no significance in LAMMPS.

7

Dear Jim,
First think about what you want to achieve? What does your rigid structure looks like? Should it move as a whole or not?

If not, then don’t define them in fix rigid. Two options

  1. do not not put them in integration
  2. Id you do, set the initial velocities = 0 and setforce = 0
8

Fix ridid does integrate and move its atoms.
Fix npt dilate partial will not move atoms not
in the fix npt group, else they will move as the
box expands/contracts.

Steve

2010/6/17 Jim Jim <[email protected]...>: