1. Jun 16, 2012

### philrainey

I have always thought that a radiation heat pump could be made if one could bend radiation perhaps beating carnots law. Do you think that if you have a big sphere at a temperture to emit visible light that radiation travelling through a vaccuum inside the sphere could be bent around a smaller sphere in the the middle perhaps by fibre optic nerves so the radiation is deflected and reabsorbted by itself as it hit the other side of the big sphere. But have it so radiation from the little sphere is able to escape to the big sphere creating a heat pump from the colder inner sphere to the warmer outer sphere. Is there any other way to bend electro magnet radiation?

2. Jun 30, 2012

### philrainey

I say this because I was reading about work people are doing to channel solar radiation from a large surface into a much smaller solar panel. They do this because large solar panels are expensive to make. If they can channel solar radiation into a small panel surely it is possible to channel similar radiation from a surface at say the same temperature as the sun surface around a smaller sphere that is say maybe 500 degrees Kelvin less. If less radiation is received by the smaller sphere than the smaller sphere radiates to the bigger. Then as there is no other heat transfer between them (because of the vacuum) and if there is no heat transfer to the outside world (just imagine there is a perfect vacuum outside the big sphere and then a perfect mirror (facing inwards) sphere. Then the smaller sphere must surely get colder and the outer one warmer. What else can happen?

3. Jun 30, 2012

### Q-reeus

You might like to read these: http://arxiv.org/abs/0910.2159v2 http://www.newscientist.com/article/dn17980-first-black-hole-for-light-created-on-earth.html
Many years ago I came up with an equivalent, but in an all-optical planar geometry (still have NDA's dating from 96 to prove priority of concept). It is imo amazingly easy to show how 2nd law can [STRIKE]be[/STRIKE] at least certainly seem to be violated (expecting flack on that one!). I may get round to posting on it reasonably soon but not in Quantum section where this thread does not really belong, and expect admin to soon move it on to elsewhere.

[Just to clarify a bit. My 'equivalent' to that given in links above was, apart from planar geometry, different in one other key respect. Instead of just an absorbing surface at final destination (cylinder core in their model), replace with an ideal Lambertian diffuse scattering surface - perfect reflection but randomized direction following Cosine rule. Time reversibility of energy flow is initially disrupted, and final equilibrium situation with equal incoming/outgoing energy flux has energy density concentrating closer to the lambertian surface. Gradient in refractive index is the other ingredient in accomplishing that (Strictly speaking, no actual gradient is needed - but things are far more efficient with either continuous gradient or multi-layer arrangement). To now create a continual disequilibrium, insert into that region say one end of a tiny thermocouple - the other end sensing the ambient environment of lower radiant energy density. Electrical power can thereby be continuously generated - at the expense of cooling the environmental surrounds. A rough outline, and of course one must take on board an overwhelming consensus opinion that some 'subtle flaw' has been overlooked. And supposing no such flaw exists, not exactly very practical for a number of reasons. But then we are here talking principle not practicality.]

Last edited: Jun 30, 2012
4. Jul 11, 2012

### philrainey

perhaps a maxium minium pattern caused by Interference could be used to avoid radiation hitting the cold heat sink (sphere that receives little radiation energy from the other) I'm not sure what kind of shape could to this but if it could be done by Interference of the radiation waves perhaps in could be done with lower frequency radiation like infer red.

5. Jul 11, 2012

### Q-reeus

Cannot fathom why this thread has not been moved to either classical or general section, but if you wish to get possibly better feedback, looks like your best shot would be repackaging the query and posting in one of those two areas - Classical Physics probably most suitable.