Re: dW turns into dQ

[Joel Rauber <Joel_Rauber@SDSTATE.EDU>]

> You misunderstand (ie., I was not clear):
> In accordance with Jim's query, I was envisioning a problem
> which involved
> applying dE= dW + dQ to a real irreversible process ( to answer some
> question?), and then in addition asked for the entropy change.

You have made your point clearer; I'm not sure where the differences lie,
but there are some, perhaps they are mostly philosophical.

> In applying the first law to the real process the aforementioned wide
> discretion is applicable in assigning dW vs dQ and thereby
> determining the
> final state properties.

I think the discretion is purely mathematical and not physical.
e.g.
10 = a+b

Mathematically there is lots of discretion as to what number to assign to
'a' and to assign to 'b'.  But I'm unwilling at this point to say that
physical considerations (the actual process that is occurring) provide no
constraints.  I think it may boil down (again radiation and convection
aside) to claiming DQ is that part of the energy transfer due to temperature
difference. DW is the other stuff.

> Then, IF YOU WISH TO CALCULATE dS
> directly from
> the relation dS=dQ/T, then you must invent a new, reversible process
> connecting the same two end states - in this calculation, dQ
> is unique to
> your chosen reversible (unreal) process.

An excellent way to calculate it.

> I added that ,since the end states are now known, dS could
> alternately be
> calculated from its (presumably known) state function.

I'm not sure of the point here, yea, if you know the state function this is
better. But if you don't know the state function, you're likely to the above
first way you mention.

> I also noted that for the (any) REVERSIBLE process dE=dW+TdS
> applies and
> dQ=TdS.   I caution that this TdS is not equal to the dQ you
> may have used
> in applying the first law to the real irreversible process (where you
> enjoyed the "wide discretion").

I'm unconvinced of the wide discretion, except in a narrow mathematical
sense you mention above.

If I may use an accounting analogy;

assets = liabilities + owner's equity

The books balance no matter how we assign numbers to the rhs; but for any
particular transaction there is a unique partitioning.

BTW, I am willing to say the partitioning may be frame dependent.

Joel