Re: partitioning dU & the first law (short)

[Brian McInnes <bmcinnes@PNC.COM.AU>]

Joel,
You wrote

> Is the partitioning of dU into DQ and DW for a specific process unique?
>
> Opinion A: yes it is unique within certain proviso's. This is my current and
> apparently lonely position.


Not that lonely, Joel.

I've been reading this discussion rather than participating (for the
most part, both you and Bob have been doing this well and have both
been spot on.)

A vital key point that seemed to be at danger of being lost in some of
the original discussions is that the entropy change for a real process
must be calculated
for an appropriate REVERSIBLE change that takes the system from its
initial
equilibrium state to its final equilibrium state.  In order to do that
calculation
we have to know the dQ for the reversible (quasi-static) path.  Not
only does this quasi static path   bear no resemblance to the path
taken in the real process, there is really no "path" for the real
process in terms of lines that can be drawn on a PV or a TS diagram.

This point is there in your postings.

John Denker raised a point about how close to complete reversibility a
process can be.  In his query he was looking at a "work" process.  As
I understand it, there isn't an "almost" reversible process.  A
process is either "reversible" or "not reversible" just like something
is either" unique" or "not unique" or "perfect" or "not perfect".  It
has always seemed to me to be one of the beauties of equilibrium
thermodynamics that its development depends so much on reversible
changes, none of which occur in nature.

Brian Mc