Re: first law of thermodynamics

["John S. Denker" <jsd@MONMOUTH.COM>]

>> aren't reversible processes necessarily near equilibrium?

I wrote:

> No.

Carl E. Mungan wrote:

> Certainly I didn't say or mean to imply that a reversible process has
> to be heat-based.

OK, we agree.

> Nevertheless I'm more than a bit surprised to hear you answer my
> initial question in the negative. I suspect it's time to back up and
> ask how you define "reversible"?

More likely we are getting tangled up in the
definition of "equilibrium".

Here's a simple example.  (I regret previous
unnecessarily-complicated examples.)  How about
a simple pendulum, or a mass on a spring.  It
converts energy from PE to KE and back, largely
reversibly, but it's nowhere near equilibrium,
according to the most-common definition of
equilibrium.  Equilibrium means "forces in balance"
and the forces aren't in balance, as we can
see from the acceleration of the mass.

This isn't worth pursuing very far;  I wasn't
trying to make a very deep point.  I violated
a rule I learned from Paul Henry:  never support
a strong argument with a weak one.  There are
a number of valid but nit-picky complaints about
the E=W+Q "law" (which I refuse to call the first
law of thermo).  These complaints should not be
allowed to distract attention from the crucial
fundamental problems with the typical ill-thought-out
W+Q approach.

Or maybe there is a deep point here:  The notion
of "equilibrium" in thermodynamics is really a
bit subtle.  In order to build a heat engine, you
need to have many, many modes that are in thermal
equilibrium while a few others (the "macroscopic
mechanical motion" modes) are not.  This separation
requires tricky engineering in practice, and is
also tricky to describe theoretically.

> I am doubtful that conversion of glucose to ATP say is reversible.

Not exactly reversible.

I am doubtful that any real-life embodiment of a
Carnot-cycle engine is exactly reversible, either.