8/10/2019 Artigo Engines Demons NatPhys2014 (1)
2/2
410 NATURE PHYSICS| VOL 10 | JUNE 2014 |
www.nature.com/naturephysics
news & views
original cycle. Following the eedback step,the traps are driven
back to the originalsingle-well configuration
(isothermalcompression) while keeping the electrostaticfield
constant. Te cycle is completed byadiabatically resetting the
electrostatic fieldto its initial value.
Te mean entropy production o theprocess is determined by
relating theobserved particle trajectories to statisticalaverages o
the thermodynamic statevariables through a suitably
adaptedcanonical fluctuation theorem8, similarto those used to
characterize the oldingenergetics in DNA-pulling
experiments9,10.Such an explicit averaging procedure isnecessary
because the one-particle systemoperates ar rom the
thermodynamiclimit, which means that the values othermodynamic
observables fluctuatesubstantially rom cycle to cycle. Te
authors
find that their process is indeed capable oextracting work rom
the thermal bath ata rate that is consistent with
theoreticallypredicted lower bounds on the
meanentropy-reduction.
At this point, it may be appropriateto add a brie technical
remark that, insimilar orm, also applies to DNA-pulling
experiments9,10. Owing to requent collisionso the trapped
colloid with the surroundingwater molecules, the experimental
set-up o Roldn et al.4is a realization o acanonical ensemble. Tus,
in principle,there is always a small probability that theBrownian
particle might spontaneously
cross the potential barrier. In this sense,the isothermal
expansion step in theirexperiments does not achieve strictsymmetry
breaking. Fortunately, however,as with all real-world
implementationso thermodynamic cycles, changes to theparameters in
the experiment occur in afinite time. Roldn and
colleagues4veriythat the mean time or the colloid to passthrough
the potential barrier (about oneweek) is much larger than the
typical cycletime (less than one minute). Tis suggeststhat
ergodicity is almost surely violatedin their case, as the Brownian
particle
does not have the time to explore the ullconfiguration
space.Regarding past and uture applications o
fluctuation theorems to complex processes,it is reassuring that
the well-controlledexperimental system o Roldn et al.4produces
results that are consistent withtheoretically predicted entropy
bounds. It is
worth noting that their study also suggeststhat memory
erasure7can be interpretedas the restoration o a broken
symmetry.Tereore, these new experiments not onlyprovide guidance or
the design o intelligentthermodynamic cycles but also elucidate
theintimate connection between inormation,
symmetry and thermodynamics.
Jrn Dunkel is at the Departmentof Mathematics,
MassachusettsInstitute of Technology, Cambridge,
Massachusetts 02139-4307, USA.e-mail: [email protected]
References1. Landauer, R. IBM J. Res. Develop.5,183191 (1961).2.
Lloyd, S. Nature406,10471054 (2000).3. Maryama, K., Nori, F. &
Vedral, V. Rev. Mod. Phys.
81,123 (2009).4. Roldn, ., Martnez, I. A., Parrondo, J. M. R.
& Petrov, D.
Nature Phys. 10,457461 (2014).5. Szilrd, L. Z. Phys.53,840856
(1929).
6. elegdi, V. L. Phys. Today53,
2528 (2000).7. Brut, A. et al.Nature483,187190 (2012).8.
Campisi, M., Hnggi, P. & alkner, P. Rev. Mod. Phys.
83,771791 (2011).9. Hummer, G. & Szabo, A. Proc. Natl Acad.
Sci.USA
97,36583661 (2001).10. Liphardt, J. Nature Phys.8,638639
(2012).11. Magnasco, M. O. Europhys. Lett.33,583588 (2012).
Published online: 28 April 2014
FRICTION
Let it slipFriction involves a complex set of phenomena spanning
a large range of length scales, but experiments assessing
the evolution of the slip-front between two dry sliding bodies
now reveal that slip can be reasonably well described
by linear fracture mechanics theory.
Robert W. Carpick and Roland Bennewitz
How do things slip? Scientists,rom Leonardo da
Vinci1toPierre-Gilles de Gennes2, have
pondered riction and slip, but thedifficulty in observing and
measuringthe behaviour hidden at the interacebetween contacting
materials remains avexing challenge. Yet the stakes are highas
riction has costly and sometimes evenlie-threatening consequences:
predictingearthquakes around geological aults(Fig. 1) and the
optimal seismic design obuildings requires knowledge o
rictionalslip dynamics across multiple lengthscales. Moreover,
rictional behaviour candetermine the efficiency and reliability
osliding mechanical elements in systemsranging rom wind turbines to
deployable
satellite components, and rictiondetermines unction in natural
systems,rom climbing geckos to the health and
strength o hips and knees.As they report in Nature, Ilya
Svetlizkyand Jay Fineberg3reveal how rictional slipis closely
related to racture. Tey addresstwo key issues that are particularly
relevantto sliding in dry conditions: measuringthe local deormation
and dynamics o themoving ront o a buried sliding interace,and
subsequently, whether one can usethe simple and well-developed
theory olinear elastic racture mechanics to modelthe results. Up to
certain limits, they findremarkable agreement between the
theory(developed or shear-loaded cracks in anotherwise uniorm
single material) and the
experiments (involving two distinct blocksin contact sliding
relative to one another).
For two materials in contact such as
the block-on-a-plane system examinedby the authors, an applied
shear orceinitially causes no relative motion, just aslight
deormation o the two materials.However, as the applied orce is
increased,slip eventually occurs. In many dry systems(that is,
systems with no lubricant), slipdoes not occur all at once but
rather bythe propagation o a pulse that originatesat the back-end o
the block, causing acrack to appear at the interace. As theleading
edge o the pulse moves orward,the slipped region grows the
crackextends until the entire block has movedorward. Te challenge
is to determine the
2014 Macmillan Publishers Limited. All rights reserved
mailto:[email protected]:[email protected]
