Energy conservation constraint on the thermal expansion coefficient

[Dave Long]

> It would seem that conservation of energy would constrain
> the [material properties]; if they could be anything you 
> wanted, you could create and destroy energy ...  I wonder 
> if real materials' coefficients are anywhere near this limit.

Do gases count as materials?  Since they seem to be the working
fluid of choice for engines, I'd guess they'd be the closest of
the cheap possibilities.

The relevant limit seems to be Carnot efficiency: (Th - Tl)/Th
gives the largest fraction of work we could hope to get out of
an engine running between a heat differential of temperatures
Th{igh} and Tl{ow}.  (degrees Kelvin, of course)

If we're only talking about a one-shot heating, then perhaps
it would be possible to get more work out of a material.  The
problem is that heat flows accomplish a transfer of energy via
an increase in entropy, but extracting work is only a matter
of energy, so in continuous operation one needs to balance the
entropy budget with a waste heat flow, therefore reducing the
amount of energy available for work.  Remove that constraint,
and one might have an energy-bound, rather than entropy-bound,
process.

In trainspotting, the "big iron" changes energy-bound problems
(accelerate this train) to entropy-bound ones (extract work
from steam/diesel combustion).  In computers, the "big iron"
changes compute-bound problems (scads of FMACS) to i/o-bound
ones (oh, you needed the results stored?).

-Dave

(wondering if dry ice is suitable for home
studio explosive molding of polyethylenes...)