# Can anyone prove this overstated-but-almost-never-justified fact from thermodynamics?

Clausius inequality states that $$\oint {\delta Q\over T}$$ equals zero for a system undergoing a reversible cycle, whereas it can’t be greater than zero for an irreversible cycle.

But everywhere, I see people and resources mentioning that “it follows” that for irreversible, it is strictly negative. HOW??

Mostly, they mention that since entropy changes for irreversible processes are positive, the above follows, except that this follows from the above once the above is justified! Else, give me an independent justification of this argument!

Otherwise, it is mentioned that irreversible processes involve dissipation, so entropy generation. But wait! An irreversible process doesn’t have to be dissipative. Say, irreversibility is due to non-quasi-staticity. Now what?

Similar is the case with the efficiencies of irreversible engines working between two temperatures.

Carnot’s theorem merely says that an engine working between two temperatures can’t be more efficient than a reversible engine working between the same temperatures.

I know that this implies that all reversible engines working between these temperatures have the same efficiency. But HOW does this imply that an irreversible engine between the same two temperatures will have a strictly less efficiency? Still people and resources blatantly mention this without any justification.

• What do you mean by "non-quasi-staticity"? That's not a term I am familiar with... Can you give an example of a process with that property? – tpg2114 Dec 7 '19 at 16:55
• A process that isn’t quasi static. – Atom Dec 7 '19 at 16:56
• Okay -- and so in a process that happens really quickly such that there are non-equilibrium effects, how do you define all of those macroscopic properties like energy and temperature? What would the integral look like for those types of problems? Does that integral work for processes that aren't in thermal equilibrium? – tpg2114 Dec 7 '19 at 17:01
• Exactly! Why then there’s so much mention of the integral being negative for irreversible cycles? – Atom Dec 7 '19 at 17:04
• @tpg2114 Actually, the integral also makes sense for the non-quasistatic processes since the temperature is not of the entire system, but the boundary at which the heat is being exchanged. And this is obviously well-defined. – Atom Dec 7 '19 at 18:26

When a process in a cycle is irreversible, it generates entropy. In order to complete a cycle all thermodynamic properties, including entropy, must be returned to their original state. That requires the system to get rid of the entropy generated by transferring it to the surroundings. The only way to transfer entropy to the surroundings is by heat. Heat out of the system is negative. Ergo $$\oint {\delta Q\over T}<0$$ for an irreversible cycle.