fix move rotate OR displace_atoms rotate for a twisting simulation?

a twisting simulation?

Dear Axel,

Thank you very much for this.

Could we achieve the same result by using many displace_atoms steps with
a time period for adjustment after each increment of pre-simulation
rotation (I imagine too much rotation cannot be applied at once -why is
that?).

technically yes, but remember that you are discretizing a continuous
motion and the larger the steps, the larger the error. the atoms that
you move with displace atoms or fix move will just go where you put
them, but the *other atoms* have to follow their motion via regular
numerical time integration.

Or else we could perform the whole rotation in a single fix move which
takes place throughout the simulation.

you can do that as well. this is what i personally would prefer,
because it has the smaller error. you will *still* need to carefully
choose the rotational velocity (and/or the time step) so that your
time integration does not diverge.

axel.

Dear Axel,

Thank you very much.
Would you have any advice for choosing the timestep, simulation time,
and period time for the move rotate fix?

Daniel

Dear Axel,

Thank you very much.
Would you have any advice for choosing the timestep, simulation time,
and period time for the move rotate fix?

pick it so it is not too small (or else the simulation will take too
long) and not too large (or else you have too large errors).

axel.