virgil1612 said:
Ken, I would like to ask you something. It's about the thread where you wonder how come, in various places, there is the wrong information that the inside pressure and density would be greater for heavier stars, when in fact it's the opposite, and the interesting discussion about the nuclear reactions - luminosity causality relation. For some reason I cannot reply in that thread.
My question is, when we have a massive star that reaches the Eddington limit and basically disintegrates, I understand that it's the radiation pressure that balances gas pressure. Does this happen roughly in the whole volume of the star or just in the outer layers?
Thanks, Virgil.
The Eddington limit applies everywhere in the star, but note a couple important points. The limit is a limit on the luminosity, given the mass, and the luminosity means just the energy that is carried radiatively (it doesn't count convection), and the mass is just the mass interior to the point in question. Also, the limit includes a reference to the opacity, which often can be a known value (when it's from free electrons), but other times can be significantly enhanced by other types of opacity. Finally, it is not completely obvious that this limit is fundamental, because a star can have various ways of exceeding the Eddington limit, at least temporarily.
So having said all that, just what the "Eddington limit" is can be a bit subtle when applied to the interior of the star! The simplest version of it is when it is applied at the surface, because then the energy flux has to be all radiative, the mass has to be all the mass of the star, and most stars that approach this limit have all their surface opacity being from free electrons. But it is still important that the luminosity limit doesn't
only apply at the surface, because we might imagine a star that can tolerate being out of hydrostatic equilibrium at its surface only, but it's a much tougher state of affairs if it is out of equilibrium throughout the star. So because it applies, in some sense, over the whole star, we tend to think of it as a limit on the maximum luminosity that the star can support, given its mass. (Note that it is sometimes turned around and described as a limit on the maximum mass of the star, given its mass-luminosity relationship, but that's wrong, because the mass-luminosity relationship would simply adjust such that any mass would be possible-- it's not the Eddington limit, but rather stability considerations that are not well known, that set the limit on how much mass a star can have.)
ETA: Second thoughts: The above subtleties are the reason we cannot be sure from theoretical grounds that stars cannot have a luminosity above the Eddington limit. If it happens in the deep interior, the star will just go convective and solve its problem, and indeed high-mass stars are highly convective. If it happens at the surface, the star can just go out of hydrostatic equilibrium, and have a strong wind, and indeed high-mass stars do have strong winds (though the energy losses to the wind can cause the luminosity seen at large distances to be reduced compared to the luminosity at the surface). Finally, the limit is derived in spherical symmetry, and stars can find various ways of breaking spherical symmetry, and indeed mass loss from very high-mass stars can often look non-spherically-symmetric. So on balance, I would say the Eddington limit is more of a benchmark than it is an absolute physical limit!