Experiment where Second Law is violated

[Metaleer]

Hi, all.

A while ago, I was told of an experiment where the Second Law of Thermodynamics was violated. The experiment involved mirrors or some other sort of optical devices to "move around" some particles, and upon moving these around, the entropy of the universe decreases.

I was wondering if anyone has some references, perhaps the name of the experiment?

Thanks.

 

[nickthrop101]

that is incorrect. It is not impossible for entropy to decrease, just highly unlikely.
It is possible for a smashed glss to suddenly come together likebefore it was smashed.
The science is if, as entropy increases in the universe, the course graining region grows, but the entropy increases becouse of the course graining region, not vice-verca, but is is possible for the course graining region to decrease in size, meaing a decreas ein entropy. So it doesn't disprove the second law :D

 

[lalbatros]

Metaleer,

This paper describes violations of the second law for a mesoscopic system of colloidal particles captured in an optical trap:

http://dspace-sub.anu.edu.au:8080/jspui/bitstream/10440/854/1/Wang_Experimental2002.pdf

Note that this does not contradict the second law since the second law cannot be contradicted, by definition.

 

[Metaleer]

nickthrop101, lalbatros, as the Second Law is taught in Classical Thermodynamics, without any prior knowledge of Statistical Mechanics, it is presented as an absolute law, which cannot be violated. It is when one studies Statistical Mechanics that one sees that indeed, it is highly unlikely for the entropy of the universe to decrease, but it is indeed possible, as the article lalbatros has linked to shows. So, understand me when I say "violation of the Second Law", I am well aware that violations are possible within the realm of systems where the number of particles is small - just wanted a rigorous article where this violation is shown to be evident and calculated explicitly.

Thanks for the link, lalbatros, I'll be sure to remember it. ;)

Thanks, everyone.

 

[marcusl]

You can Google the "Fluctuation Theorem", which discusses the conditions under which temporary violations of the 2nd Law can appear. Note that these "violations" are microscopic and fleeting, and are quickly reversed, leading to overall macroscopic conservation and validity the 2nd law.

 

[Vanadium 50]

I would argue that entropy is only defined for large enough N such that you don't see violations of the 2nd law. If your system is small enough that these fluctuations are important, it's too small to describe accurately with macroscopic quantities like entropy.

 

[Curl]

I would argue that entropy is only defined for large enough N such that you don't see violations of the 2nd law. If your system is small enough that these fluctuations are important, it's too small to describe accurately with macroscopic quantities like entropy.

Yeah, these papers on "I VIOLATED 2ND LAW!11!1!" are just trying to get attention. Nobody cares about the small-system short-timescale experiments which, unsurprisingly, can find a system out of its most probable macrostate. We knew that all along.

 

[Metaleer]

You can Google the "Fluctuation Theorem", which discusses the conditions under which temporary violations of the 2nd Law can appear. Note that these "violations" are microscopic and fleeting, and are quickly reversed, leading to overall macroscopic conservation and validity the 2nd law.

Thanks, I'll keep it in mind. :)

I would argue that entropy is only defined for large enough N such that you don't see violations of the 2nd law. If your system is small enough that these fluctuations are important, it's too small to describe accurately with macroscopic quantities like entropy.

I am talking about entropy, the Second Law, etc... as it is defined and understood in a course on Classical Thermodynamics, where there is no hypothesis on what matter is made of. Yes, of course, when one studies Statistical Mechanics, we see that fluctuations show up in small systems.

Yeah, these papers on "I VIOLATED 2ND LAW!11!1!" are just trying to get attention. Nobody cares about the small-system short-timescale experiments which, unsurprisingly, can find a system out of its most probable macrostate. We knew that all along.

I disagree, these are actual experiments to back theoretical results up. Boltzmann's equation is all nice and dandy, but seriously, has an explicit experiment ever been mentioned to you in a stat. mech class where the Second Law is violated? Of course "we knew that all along"; ever seen it actually happen? is what I'm saying.

On a sidenote, it's a good method to get some of these perpetual motion cranks to shut the heck up. They all unknowingly claim that the Second Law is something universal that can't ever be violated and that physicists are trying to cover up for the oil and gas companies as a part of some conspiracy, bla bla... no siree bob, physicists have known this for a while, and they've done experiments to prove it.

 

[Vanadium 50]

Yes, but Classical thermodynamics breaks down at small scales, just like the Classical mechanics it is based on breaks down.

 

[Curl]

I disagree, these are actual experiments to back theoretical results up. Boltzmann's equation is all nice and dandy, but seriously, has an explicit experiment ever been mentioned to you in a stat. mech class where the Second Law is violated? Of course "we knew that all along"; ever seen it actually happen? is what I'm saying.

yeah it's called a thought experiment and I've made up about 20. Nothing difficult here, its called probability.

Most books discuss this thoroughly.

 

[lalbatros]

Thermal noise in resistances can easily be heard.
It is even impossible to amplify a very small signal without discovering the thermal noise.
That's why high quality amplifier are specified by the Signal / Noise ratio.
That's why ultra-sensitive electronic equipment must be cooled.

This thermal noise in resistance is very will known and easy to observe.
It is a violation of the second law, obviously.

 

[Vanadium 50]

Your last sentence doesn't follow from your first 5.

 

[marcusl]

Thermal noise in resistances can easily be heard.
It is even impossible to amplify a very small signal without discovering the thermal noise.
That's why high quality amplifier are specified by the Signal / Noise ratio.
That's why ultra-sensitive electronic equipment must be cooled.

This thermal noise in resistance is very will known and easy to observe.
It is a violation of the second law, obviously.

Not at all obvious. In fact, Nyquist's theoretical derivation and explanation of thermal noise, published in 1928 shortly after Johnson's initial observations, used the 2nd law of thermodynamics as the starting point. His paper is beautifully written, by the way. Here is the citation:

http://scholar.google.com/scholar?q=h+nyquist+thermal+agitation&hl=en&btnG=Search&as_sdt=1,6"