[Phys-l] Reversible versus quasi-static processes (was Re: PV question)

[John Mallinckrodt <ajm@csupomona.edu>]

Leigh Palmer wrote:

> I cannot follow this "thread". The metaphor is inappropriate for  
> such a tangled mess. I want to make some disjoint comments that may  
> or may not answer some of the questions that have arisen during  
> this discussion. I hope that thinking about these will help dispel  
> misconceptions and misuse of conventional terminology.
>
> [perfectly reasonable commentary omitted]
>
> I just did that to assure myself that I still knew my classical  
> thermodynamics. If I've said anything that is incorrect I would  
> appreciate correction.

In my judgment Leigh still knows his classical thermo and I certainly  
agree with him that this thread moved well away from the simple  
original question almost immediately.  (Leigh may have been gone too  
long to remember that it's always like that around here!)

With regard to "reversibility," it's clear that there is a lot of  
confusion "out there" about what constitutes a reversible process and  
what the difference is between a quasi-static and reversible  
processes.  For instance, Wikipedia, defines a reversible process as  
"a process that can be "reversed" by means of infinitesimal changes  
in some property of the system without loss or dissipation of energy."

   http://en.wikipedia.org/wiki/Reversible_process_(thermodynamics)

I take issue with this definition because it suggests that there are  
processes in which energy is "lost."

The Wikipedia article also provides an alternate definition: "a  
process that, after it has taken place, can be reversed and causes no  
change in either the system or its surroundings."

I'd be more comfortable with this definition if it read, "a process  
that could, in principle, be reversed leaving both the system and its  
surroundings in their initial states."  But isn't it simpler and less  
ambiguous to define a reversible process as "a process that does not  
increase the total entropy of the system and its surroundings" or  
simply "a process that does not produce new entropy"?  Personally, I  
prefer the even simpler, "a dissipationless process," as it focuses  
attention on the real issue--dissipation, which is THE mechanism by  
which new entropy is ALWAYS produced.

Unfortunately, Wikipedia muddies the water in its article on  
quasistatic processes by saying, "A quasistatic process often ensures  
that the system will go through a sequence of states that are  
infinitesimally close to equilibrium, in which case the process is  
typically reversible."  Even if this were true, the statement is  
rendered essentially meaningless by the weasel words "often" and  
"typically" and it simply is not the case that a quasistatic process  
is "typically" reversible.

It seems to me that the confusion arises as a result of a very common  
failure to be clear about 1) the central role of dissipation and  
entropy production and 2) the system to be considered.  I suspect  
there may be no universal agreement on definitions, but I'll propose  
two for target practice:

QUASISTATIC PROCESS

Definition: A system undergoes a quasistatic process if and only if  
no dissipation (or entropy production) occurs within the system.

Notes: The only way to do this is to ensure that the system's  
mechanical constraints (including things like ambient E, B, and g  
fields) change slowly AND that its physical boundaries are always  
held at a temperature that is close to that of the system itself.   
What is "slow" enough and "close" enough depends on how picky one  
wants to get about the absence of dissipation or the amount of new  
entropy that one is willing to tolerate.  Any significant friction in  
a moving piston would disqualify the process IF the system under  
consideration includes the piston and/or container walls, but is  
completely irrelevant if the system under consideration is a gas  
confined to the container.  In any event, note that the SYSTEM must  
be unambiguously defined.

REVERSIBLE PROCESS

Definition: A process is reversible if and only if no dissipation (or  
entropy production) occurs due to the process.

Notes:  Implicit in the definition is that "the system" is  
"everything there is," i.e., "the universe."   Thus, our attention is  
focused here only on the PROCESS.  We do not question the  
reversibility of a Carnot cycle simply because a free expansion might  
be happening somewhere else in the universe at the same time.  This  
is NOT because the free expansion isn't taking place within "the  
system;" it is because the free expansion isn't part of the PROCESS  
we are considering.  If a process includes moving a frictional  
piston, then the process is not reversible.  No real process is  
reversible, so the discussion of reversibility is always about  
principles, not experimental procedures or apparatus.

Fire away!

John Mallinckrodt
Cal Poly Pomona