problem with smd, periodic boundary, and negative velocities


I’ve encountered this weird behavior of the smd fix. I want to drag a group of atoms, which I have named “puller”, along -Y (negative velocity). It happens that when the puller reaches the Ymin box boundary it stops instead of crossing it as PBC would impose. Strangely, if I do the same along +Y (positive velocity), it works as expected: it crosses the Ymax boundary and reappears on Ymin and keeps going.

These are the main lines:

boundary f p f
group puller id 3715 3717 3719 3721 3723 3725 3727
fix 5 puller smd cvel 0.001 -0.01 tether NULL 0.0 NULL 0.0

It smells like a bug…