🚀 go-pugleaf

http://focus.aps.org/story/v23/st10
"In collisions between two deformable or sticky objects, some or all
of the initial energy is usually lost to heat. But computer
simulations of nanocluster collisions, as described in the March
Physical Review E, show that thermal fluctuations in the tiny
projectiles can cause them to rebound with more kinetic energy than
they started with. This energy boost would seem to violate the second
law of thermodynamics, which bans heat from being turned directly into
energy of motion. (...) The fraction of these anomalous rebounds
increased with the cluster temperature. "When the temperature
increases, more and more vibrating modes are excited," Kuninaka says.
These vibrations can sometimes give an extra kick to the collision,
like a gymnast pushing off the pummel horse to get a greater lift. The
transfer of energy from random internal motion of the atoms to the
movement of the whole cluster creates a decrease in entropy and
therefore an apparent violation of the second law of thermodynamics.
But that's OK at small scales."

That's OK at large scales as well, as demonstrated by macroscopic
contractile polymers:

http://pubs.acs.org/doi/abs/10.1021/jp972167t
J. Phys. Chem. B, 1997, 101 (51), pp 11007–11028
Dan W. Urry, "Physical Chemistry of Biological Free Energy
Transduction As Demonstrated by Elastic Protein-Based Polymers"

As far as I know, the only paper explaining (or trying to explain) how
biased thermal motion can phenomenologically be expressed as violation
of the second law of thermodynamics is this one:

http://www.wbabin.net/valev/valev2.htm

Pentcho Valev
pvalev@yahoo.com