Black body radiation at atomic level I need help in black body radiation. I know quantitative equations of black body radiation. what I need to know is what is happening inside a black body at atomic level. What I mean is, if light photons is colliding with electrons, then electron is going to the next energy level and then, when come back to the original level it emits the energy. From black body theory we know that if we heat up a black-body with infrared light up to 5000k it will emit all short of visible light. That is colliding by low energy photon we get high energy photon. Doesn't it disobeys the rule that a photon's energy is completely taken by an electron while jumping to high energy level and given off totally when coming back. Then how more low energy photon can produce high energy photon? please explain.
 A: Many explanations of blackbody radiation leave out what is actually going between the blackbody itself (say the wall of an oven) and the electromagnetic field (the photons). This leads to a lot of confusion, including the confusion you are expressing now.
The important thing to note is that the whole idea of blackbody radiation relies on the ability of the walls of the blackbody to thermally equilibrate with the electromagnetic field. This means there must be some way for energy or something to transfer between the two systems (blackbody, EM field). In this instance you have correctly identified the fact that this coupling is mediated by the absorption and emission of photons by the atoms and molecules which make up the surface and bulk of the blackbody. For example, if the atoms and molecules were NOT able to interact with light then there would be NO photons, even in a very hot blackbody. Experiments have observed this effect by creating the blackbody out of some material which does not absorb or emit in a certain frequency band and you can observe dips or gaps in the blackbody spectrum of the object.
Now to more directly address your question. You seem to be concerned about conservation of energy in blackbody absorption and emission. I think the key point to realize is that the blackbody doesn't emit photons only because it absorbs photons (if it did then what you are saying would be correct.) Rather, the blackbody emits photons because
1) it has a lot of energy
1a) This energy is stored in many possible ways in the blackbody, It may be stored in chemical bonds between atoms/molecules in the blackbody, it might be stored as electrons in excited states, it may be stored as mechanical soundwaves (phonons) in the material etc.
2) That energy is distributed in such a way that the blackbody has some temperature T
3) The blackbody is coupled to the electromagnetic field.
Since the blackbody is at some temperature, and it is coupled to the electromagnetic field, then the laws of statistical mechanics (entropy) tell us that it is most likely that the blackbody will emit and absorb radiation in such a way that after some short amount of time the two systems (EM field and blackbody) will be in thermal equilibrium. The powerful part about statistical mechanics is that you don't have to know HOW the energy redistributes itself to be able to know that it will.
Of course that last sentence in the last paragraph is really unsatisfying so I'll try to answer directly the main part of your question which I'll rephrase as: How can we get very high energy photons from a blackbody when there don't seem to be any high energy photons around at the beginning. The answer is that there are mechanics for energy to distribute in such a way that a high energy photon might appear. For example, maybe an atom can absorb multiple low energy photons as well as some phonons and collisions with other atoms etc. Occasionally an atom could come to be in a very highly excited state this way and it could somehow emit a high energy photon.
I want to say that this is also generally tied to the idea that quantizing the electromagnetic field (light comes in discrete energy packets called photons) solves the ultraviolet catastrophe. In other words, Planck's distribution captures the fact that these high energy events are less likely than the lower energy ones, but I haven't thought through it clearly and don't have time at the moment. Maybe someone else can weigh in on this.
