Can photon-photon interactions alone produce black body radiation?

[Rap]

Suppose you have a container with perfectly reflecting walls, containing electromagnetic radiation that is not in equilibrium (i.e. does not have a Planck distribution of energies.) Will photon-photon interactions (QED and/or gravitational) produce a Planckian distribution after a sufficiently long time?

I don't care how long - fifty gazillion times the age of the universe, I don't care.

 

[mathman]

I suspect that bouncing off the walls would lead to thermal equilibrium very quickly. Photon-photon interactions (except right after the big bang) are extremely rare.

 

[Rap]

I suspect that bouncing off the walls would lead to thermal equilibrium very quickly. Photon-photon interactions (except right after the big bang) are extremely rare.

No, bouncing off the walls does not change the energy of a photon. If the energy distribution were out of equilibrium, it would remain out.

 

[Vanadium 50]

Sure it will. It can even change the number of photons.

 

[Rap]

Sure it will. It can even change the number of photons.

I don't see how - the energy has to be conserved, so if one photon hits, two photons reflect, then they are both of lower energy. That's not perfectly reflecting.

 

[mathman]

No, bouncing off the walls does not change the energy of a photon. If the energy distribution were out of equilibrium, it would remain out.

Bouncing off the wall means an interaction takes place with the material in the wall. The wall is at a temperature - hot photons hitting the wall will tend to cool off, while cold photons tend to heat up.

 

[Rap]

Bouncing off the wall means an interaction takes place with the material in the wall. The wall is at a temperature - hot photons hitting the wall will tend to cool off, while cold photons tend to heat up.

But this is counter to the assumption of perfectly reflecting walls. You are changing the subject.

 

[Bill_K]

Photon-photon interactions (except right after the big bang) are extremely rare.

But they do take place. And he gave us fifty gazillion times the age of the universe, so one scatter every 13.7 billion years is still fifty gazillion collisions.

 

[mathman]

But this is counter to the assumption of perfectly reflecting walls. You are changing the subject.

There is no such thing. The walls have to be made of something. That something will have a temperature.

 

[Dickfore]

Bouncing off the wall means an interaction takes place with the material in the wall. The wall is at a temperature - hot photons hitting the wall will tend to cool off, while cold photons tend to heat up.

What is a "hot photon" and a "cold photon"?

 

[Rap]

There is no such thing. The walls have to be made of something. That something will have a temperature.

You are missing the point of the question. The reflecting walls are a way of bounding the system in order to ask a question. They are not the point of the question. The question is whether photon-photon interactions will thermalize a photon gas that is out of equilibrium.

Assume a photon gas of practically infinite extent so that boundary conditions are practically negligible - will photon-photon interactions thermalize that photon gas?

Assume a closed universe composed only of photons, so that there are no boundary conditions - will photon-photon interactions thermalize that photon gas?

Do you see what I am asking?

 

[fzero]

You are missing the point of the question. The reflecting walls are a way of bounding the system in order to ask a question. They are not the point of the question. The question is whether photon-photon interactions will thermalize a photon gas that is out of equilibrium.

Assume a photon gas of practically infinite extent so that boundary conditions are practically negligible - will photon-photon interactions thermalize that photon gas?

Assume a closed universe composed only of photons, so that there are no boundary conditions - will photon-photon interactions thermalize that photon gas?

Do you see what I am asking?

This abstract http://meetings.aps.org/Meeting/DPP11/Event/152079 would suggest that the answer is yes, which I don't think is unexpected.

 

[Rap]

This abstract http://meetings.aps.org/Meeting/DPP11/Event/152079 would suggest that the answer is yes, which I don't think is unexpected.

Ha! Thank you! This is the sort of thing I am looking for!