[gmx-users] About Pulling code in Gromacs 5.0.2

[Alex]

The artifacts warning is general (mentioned in the manual) and has nothing to do with your
system (aside from your mdp), everyone using absolute reference frame
gets this warning.

That said, the error you're getting is actually descriptive. So:

1. Check if the system is stable and happy without pulling.
2. Make sure the pulling rate is compatible with your nstlist setting.
If the neighbor list update rate can't keep up with your pulling rate,
that's a problem. High pulling rates in thermostatted systems are also
generally problematic from the physical standpoint. Severely
problematic. Steered MD is to be carefully designed.
3. Make sure your boundaries are taken care of properly, in case your
pulled object meets the wall.

Alex



hy> Hello all,

hy> Recently, I preformed a SMD, pulling a nano diamond across the membrane. I
hy> found that pull_geometry position in the Gromacs 4.6.5 (suggested in the
hy> tips of the US tutorial by Justin.A ) was replaced by pulling with absolute
hy> reference point in Gromacs 5.0.2. After defined pulling parameters
hy> according to the manual, Gromacs gave me a warning that "This will lead to
hy> artifacts". I really want to know what did it mean, so I ignored it by
hy> using maxwarn 1 and saw what would happen. Finally I got a segment fault
hy> that a bonded atom moved away, out of the cut-off distance.

hy> Can you give me some advises? How can I treat the pulling correctly for
hy> this penetrating process and why does the absolute reference lead to
hy> artifacts.

hy> Many thanks!

hy> Hang