Freezing atom in place.

Hello,

I am trying to simulate a nano structure. I want to freeze a few of the atoms in the simulation. I have noticed that regardless of all my attempts, the atoms I freeze continue to move. I have separated the atoms into two groups. Either unfrozen, or freeze. The program seems to be recognizing the groups just fine.

I set the frozen atoms velocities to 0.0 and this is verified in my dump file. I see the velocities staying at 0.0 at all times. However the atoms are still managing to move despite having no velocity. I did not include the atoms in the NPT integration and I also told the frozen atoms to have their forces continuously set to 0.0.

What am I misunderstanding ? How can something move with no velocity?

Thanks
Tim

Timestep 0 for Atom 434 (x, y, z, vx, vy, vz)
71.4818 76.7641 45.7597 0 0 0

Timestep 50,0000 for Atom 434 (x, y, z, vx, vy, vz)
434 71.4999 76.7836 45.7713 0 0 0

Timestep 100,000 for Atom 434 (x,y , z, vx, vy, vz)
434 72.5545 77.9161 46.4464 0 0 0

Hello,

I am trying to simulate a nano structure. I want to freeze a few of the
atoms in the simulation. I have noticed that regardless of all my attempts,
the atoms I freeze continue to move. I have separated the atoms into two
groups. Either unfrozen, or freeze. The program seems to be recognizing the
groups just fine.

I set the frozen atoms velocities to 0.0 and this is verified in my dump
file. I see the velocities staying at 0.0 at all times. However the atoms
are still managing to move despite having no velocity. I did not include the
atoms in the NPT integration and I also told the frozen atoms to have their
forces continuously set to 0.0.

What am I misunderstanding ? How can something move with no velocity?

because you change the box. this is usually done by transforming all
coordinates to fractional coordinates in the coordinate system of the
cell vectors, then adjusting the box vectors and recomputing the
absolute positions from them. please see the fix npt documentation for
details. search for the "dilate" flag. please very carefully check if
keeping immobilized atoms on fixed absolute coordinates has negative
side effects. it is best when those are located as close as possible
to the point around which the remaining coordinates are scaled.

axel.