Sure. I'll go ahead and attach the input file.
please always reply to the mailing list not just
to individual people. we want the resolution of
any question and problem to be recorded in
the mailing list archives, so that others can
benefit from it in the future. thanks.
now on to your question.
the main explanation for your "stuck" atoms
is this line in your input:
fix 1 all nvt temp 300.0 300.0 0.1
in combination with using fix setforce to
move your atoms boundary atoms. this
force is rather large and will generate
significant kinetic energy which drives
the atoms in the way you want, but also
results that the nvt thermostat is actually
*cooling* all the atoms. i believe a better
and more realistic setup would be to
use fix move to move your upper and
lower boundary atoms and then do a
fix nve integration on the rest after an
initial equilibration at the desired temperature.
due to the very high frequency of your
motion, you still have to expect a significant
amount of kinetic energy (= temperature)
being generated, but i'd consider that quite
to be expected and realistic, for as long as
you have a large enough "buffer" between the
tips and boundary atoms. thermal dissipation
along those is usually quite slow.
axel.