Dear all,
Is there any chance that setting an aveforce fix in a certain
direction ,which is also thermostated via langevin + nve or nvt, can
affect the equations of motion in a way that mess up the dynamics? It is
quite generalistic the question, but I am experiencing some issues in
this regard which are solved if I just use addforce. I tried to
understand if that could happen by looking at the source code but it is
not clear to me at a first glance. If just someone could confirm me that
there could be issues due to how the step loop works, would be nice.
I would expect that once the forces are calculated derived from all the
contributions of different potentials (bond, pair, coul,...) the
aveforce fix average this forces in a certain direction and then set
(fave + fextra) as the new force acting on an atom. Then the bits
related to the particular thermostat are added to the equations of
motion and integration takes place (talking roughly).
BTW, I am calling in this order for both Langevin and Nose-Hoover
thermostating cases:
fix aveforce
fix langevin
fix nve
fix aveforce
fix nvt
Thanks in advance,
Eduardo