[lammps-users] Freeze atoms during minimization

I'm setting up a dislocation. I would like to displace some of the atoms according to linear elastic theory, then freeze those, and minimize the others nearer to the dislocation core by conjugate gradient. I understand that during a normal run, I can just set a fix nve on just the atoms near the core; this will freeze the ones I've displaced by hand. However, the minimize command doesn't seem to have a similar dependence on region - it just minimizes the entire system. Is there a way around this? I'm not sure if maybe fix freeze would work, but I don't have a granular system.


Hi Erin,

What you want to do is designate those atoms you've displaced due to linear
elasticity as their own group, and then use a "setforce" fix to set the
forces on those atoms to have zero force components before you do the
minimize. This fix will be in place during the minimize and those atoms will
not move.

Jon Zimmerman