Re: [Phys-l] T dS versus dQ

[John Denker <jsd@av8n.com>]

On 01/12/2010 10:30 PM, Bernard Cleyet wrote:
> ... I assumed insulated, isolated, and impermeable to
> everything (except the piston) are identical.  Evidently not.

They're not strictly identical, but close enough
for present purposes.  In the present context they 
have the same consequences.

The more important distinction is that "adiabatic" (in 
the sense of thermally insulated, i.e. no entropy flow 
across the boundary of the region) is not equivalent 
to "adiabatic" (in the sense of isentropic, i.e. no 
change in entropy content of the region).

Just because no entropy flowed across any part of
the boundary (including the piston) does *not* mean
that there was no change in entropy content.  Entropy
is not a conserved quantity!  Models that require
transfers to/from a fake entropy reservoir are not
good models.

To say the same thing in other words:
 -- stirring does not transfer entropy across the
  boundary
 -- stirring causes entropy to be created from
  scratch within the fluid.

To say the same thing yet again:  If we were dealing
with a conserved quantity such as energy, then 
no energy was transferred across the boundary
would be equivalent to saying there was no change
in the energy content of the region ... but that's
not what we are talking about.

Entropy is not a conserved quantity.

Irreversible processes create new entropy from scratch.

Irreversible processes are common in the real world.

A model of thermodynamics that cannot properly handle 
irreversible processes is going to give the wrong
answer when applied to Carl's scenario.

Entropy is not a conserved quantity.