Ian wrote:
> A refrigerator is a device which removes heat from one location, at
> the cost of generating additional waste heat at various locations.
> The system as a whole grows hotter (the system includes not only the
> refrigerator but the power plant generating electricity for it).
And the huge heat sink of the enviroment outside (ie- a materially
closed system of powerplant + refridgerator can still dump heat outside
- think of a refridgerator powered by a battery for instance). The point
is that the powerplant + refridgerator can indeed cool itself, according
to the laws of thermodynmaics.
>>> This involves elementary thermodynamics. Entropy increases.
>>> Concentrating heat in one spot and direction...
>> ...is done all the time and...
> Only for _part_ of a system. You can't charge a laser beam with
> waste heat and use that to cool the system, it's in violation of
> thermodynamics. You can't take waste heat, which is high entropy,
> and send it off in a laser beam, which is low entropy, without
> generating even more waste heat.
Here's my conceptual problem: imagine a black box (how most of my
thermo was taught to me, and obviously about how good a grasp I have on
it now ;-). You say I can't take waste heat and pump it into a low
entropy form (like a laser) *without generating even more waste heat*.
In other words globally entropy must increase. I agree.
But you can arrange (or at least imagine) a system that does this in
a local sense (refridgeration). The catch is that you need an outside
source to power it (ie- waste heat cannot be the only think powering
it), and obviously since net entropy increases, if I'm lowering the
internal entropy of the system (temperature) I must be increasing the
external entropy (since this is the "temperature" of the laser (low, in
a thermodynamic sense), I need to dump a lot of photons per second (a
*lot*!) - quite an engineering feat, but not a violation of
thermodynamics).
See my problem? It's quite easy to see how to cool something, and a
laser represents one way to pump entropy around (a terribly inefficient
one? Certainly - but one that can work against a huge temperature
gradient).
> If waste heat is leaving a spaceship, it _must_ be in a high entropy
> form.
Maybe, but you can't prove that by a simple application of
therodynamic law. Afterall, there are chemical reactions that lower
their temperature, but increase the (global) entropy, a similar
situation.
Sorry, I'm just not getting it. As an engineering problem I see it,
but not as a consequence of the laws of thermodynamics.
--
Brian Davis