Why does gravity collapse and stars explode?

[Josh S Thompson]

Why does everyone say gravity collapses and the star explodes?

wouldnt the constant tug of gravity never allow this to happen.

The fuel burns out and the star collapses
why isn't it
the fuel is burns out little by little and gravity condenses the star little by little as the fuel burns out.

 

[PeterDonis]

why isn't it
the fuel is burns out little by little and gravity condenses the star little by little as the fuel burns out.

Because that's not what happens. At least, not in a supernova, which I assume is what you're talking about here. The fuel does not burn out little by little; that's the whole point. When a star goes supernova, it's because the rate of nuclear reactions in the core drops quickly, which means the pressure of the core also drops very quickly, and the core collapses quickly. When the star is under normal conditions, the rate of nuclear reactions is constant; it doesn't decrease little by little, any more than the rate at which your car's engine burns fuel decreases little by little as your gas tank empties.

 

[Josh S Thompson]

Because that's not what happens. At least, not in a supernova, which I assume is what you're talking about here. The fuel does not burn out little by little; that's the whole point. When a star goes supernova, it's because the rate of nuclear reactions in the core drops quickly, which means the pressure of the core also drops very quickly, and the core collapses quickly. When the star is under normal conditions, the rate of nuclear reactions is constant; it doesn't decrease little by little, any more than the rate at which your car's engine burns fuel decreases little by little as your gas tank empties.

Why does the rate of nuclear reactions drop quickly?

Yes, I'm assuming the rate of nuclear reations is constant and I'm assuming the outward pressure would decrease at a constant rate.

So what you saying; since the outward pressure decrease, the gravity condenses the star more making the nuclear reations go faster?

 

[PeterDonis]

Why does the rate of nuclear reactions drop quickly?

As I understand it, it's because the buildup of reaction products in the core affects the reaction rate, and the effect is highly nonlinear. (More on this in a follow-up post.)

I'm assuming the rate of nuclear reations is constant and I'm assuming the outward pressure would decrease at a constant rate.

These two assumptions are inconsistent. If the reaction rate is constant, the outward pressure will be constant. (This is basically the state during the star's normal lifetime.) If the pressure is decreasing at a constant rate, the reaction rate must be decreasing at a constant rate. (That is not what actually happens when the star collapses, as I said above.)

 

[PeterDonis]

since the outward pressure decrease, the gravity condenses the star more making the nuclear reations go faster?

Not the same reactions, no. A star that undergoes a typical Type II supernova has a core in which silicon is fusing to iron. But outside the core, there are layers like the layers of an onion, in which progressively lighter elements are present, undergoing their own fusion reactions--key ones are oxygen, carbon, helium, and (in the outermost layer) hydrogen.

The reason the star goes supernova is that, once iron is formed by fusion, no further energy can be gained by nuclear reactions; forming heavier nuclei by fusing iron would require an input of energy. So the iron that gradually builds up in the core is basically useless for producing pressure to hold the star up against gravity. Once enough iron builds up, the silicon fusion reaction rate begins to drop precipitously; this happened in the previous fusion stages too (oxygen, carbon, helium, hydrogen), but in those previous stages, there were further fusion reactions that could start up to compensate. With iron, as above, there are no further fusion reactions possible, so when the silicon fusion reaction rate starts to drop precipitously, there is nothing else that can compensate. So the core pressure drops precipitously, and the core collapses.

The collapse of the core releases a large amount of energy. This energy gets deposited in the outer layers of the star and initiates explosive nuclear reactions. The energy released is more than enough to overcome gravity, and all the rest of the star except the collapsed core flies away. The core itself undergoes neutron production reactions, turning the iron into neutronium (and possibly producing even more exotic states of quark matter); whether this allows the core to stabilize as a neutron star or not depends on the details of the specific case.

 

[rootone]

All stars begin their life as mainly hydrogen which undergoes fusion turning ultimately into helium.
The fusion releases a lot of energy in various forms and this causes outward pressure from the core which balances the inward directed pressure of gravity.
Most stars are small and never get beyond that stage, (red dwarfs), they just slowly fuse the hydrogen over a very long time and when the hydrogen is exhausted nothing more will happen. The star will be able to maintain it's internal heat for a long time will eventually become a dark mass of mostly nothing but helium.
Larger stars such as our Sun will go through other stages of fusion, burning helium to produce elements such as carbon and oxygen, they don't explode but often shed their outer layers in a slower process than exploding. The eventually inert core is left very slowly cooling after that, this is a white dwarf star.
Larger still stars can fuse elements up to iron and nickel, and those are in danger of exploding as a supernova.
A supernova can also be caused when a stable white dwarf star is in a binary system with another star.
The small but dense dwarf star may 'steal' some the outer material from the companion star and that can lead to explosive runaway re-ignition of fusion for the dwarf star.

 

[PeterDonis]

when the hydrogen is exhausted nothing more will happen. The star will be able to maintain it's internal heat for a long time

Not very long on a cosmic timescale: only a few million years. The star will gradually shrink over that time period as it loses its heat. Eventually it will be a black dwarf (made of white dwarf matter but too cool to emit visible radiation).

Larger still stars can fuse elements up to iron and nickel, and those are in danger of exploding as a supernova.

This is a Type II supernova, the kind I was describing.

A supernova can also be caused when a stable white dwarf star is in a binary system with another star.
The small but dense dwarf star may 'steal' some the outer material from the companion star and that can lead to explosive runaway re-ignition of fusion for the dwarf star.

This is one form of Type I supernova. It is actually triggered by the mass of the white dwarf going above the Chandrasekhar limit due to accretion of matter, making the white dwarf unstable. The re-ignition of fusion happens as it collapses to either a neutron star or a black hole.

 

[Josh S Thompson]

Ok, so for type 2 supernova. You have a build up of iron core in a red giant and the fusion process slows and collapses the core into a neutron star with a violent reaction. Or is it not a red giant.

I'm not understanding the highly nonlinear nature of the rate of fusion reactions.
I also don't really understand how so much energy gets released and what type of nuclear reaction occurs.
Thank you for thorough answers it is very interesting

 

[SteamKing]

The article at this link discusses a variety of topics relating to the evolution of stars during their lifespans:

http://www.astro.caltech.edu/~george/ay20/Ay20-Lec9x.pdf

On p. 34 of the link above, a breakdown of the lifespan of a star with an initial mass 25 times that of the sun is given. At that size, the star only lasts about 8 million years from start to finish, and the last few stages of its life occur with increasing rapidity.

 

[PeterDonis]

for type 2 supernova. You have a build up of iron core in a red giant and the fusion process slows and collapses the core into a neutron star with a violent reaction. Or is it not a red giant

I don't think the star would necessarily be a red giant. The key thing is that the star has to be massive enough to have silicon fusing to iron in its core. Stars of various spectral classes can meet that requirement.

I'm not understanding the highly nonlinear nature of the rate of fusion reactions.

That's true of basically any reaction where the reaction products are confined in the reaction region; it's not even specific to nuclear reactions. The rate of the reaction basically depends on the relative concentrations of the reactants and the products, and the dependence can be highly nonlinear. So, for example, hydrogen fusing to helium exhibits the same phenomenon--five or six billion years from now, the sun's hydrogen to helium reaction rate in its core will drop off fairly rapidly when enough helium has built up.

However, at any stage of fusion prior to silicon to iron, the "reaction products" can themselves undergo further fusion reactions to yield energy. Helium can fuse to carbon and oxygen (and neon), carbon can fuse to magnesium, oxygen can fuse to silicon, and silicon can fuse to iron. So as the rate of one fusion reaction drops off, the rate of another one can increase, so energy continues to be released and there is not a sharp drop in pressure. For example, when the rate of hydrogen to helium fusion drops off in the sun's core, helium fusion to carbon and oxygen will start up and continue to supply pressure to keep the sun from collapsing. (The sun will be a red giant at this point.)

But if a star is massive enough to get to the stage where silicon is fusing to iron in its core, since iron can't produce energy by further fusion reactions, once enough iron has built up in the core and the silicon to iron fusion rate begins to drop off sharply, there is nothing to replace it, so the pressure in the core drops sharply and triggers a supernova.

The other key thing to understand is that, for a supernova to occur, the star's core must be depending on kinetic pressure (pressure due to high temperature, caused by fusion reactions releasing heat) to hold it up. That means the star's core (not the whole star, just the core) must be over the maximum mass limit for a white dwarf, the Chandrasekhar limit. If a star's core is less massive than that, it does not need kinetic pressure to hold it up, so when fusion reactions stop, it will not go supernova. For example, in the sun's core, fusion reactions are expected to stop when the core is mainly composed of carbon and oxygen (because the core temperature will never get high enough to ignite fusion of those elements into heavier elements). At that point, the sun will be a red giant, and its core will be a white dwarf, and the core will just stay a carbon-oxygen white dwarf.

I also don't really understand how so much energy gets released and what type of nuclear reaction occurs.

The star's core starts out at white dwarf density, but, as above, it needs kinetic pressure to hold it up. When the kinetic pressure goes away, the core is not stable at that density; if it can be stable at all, it can only be stable at neutron star densities. That means the core has to collapse from a size of roughly 10,000 kilometers to a size of roughly 100 kilometers. The matter of the core will gain a lot of kinetic energy in the collapse, which gets released in the supernova explosion.

The nuclear reactions that take place basically convert the core's matter from iron to neutronium, i.e., the protons in the iron nuclei combine with the electrons to form neutrons, emitting gamma rays and neutrinos. The neutrinos don't interact with the rest of the matter of the star; they just fly off at (close to) the speed of light. The gamma rays are what heat up the rest of the star's matter and blast it outward in the supernova explosion.

 

[Janus]

Ok, so for type 2 supernova. You have a build up of iron core in a red giant and the fusion process slows and collapses the core into a neutron star with a violent reaction. Or is it not a red giant.

I'm not understanding the highly nonlinear nature of the rate of fusion reactions.
I also don't really understand how so much energy gets released and what type of nuclear reaction occurs.
Thank you for thorough answers it is very interesting

It's not a slowing of the fusion process that causes the core to collapse per se, so much as it is that the fusion moves outward from the core. When the star starts fusing silicon at the core and iron builds up in the core, fusion at the core does come to a stop but fusion in the star does not. The point of silicon fusion moves outward. In fact this has happened before. When silicon started forming at the center from oxygen fusing, the oxygen fusion moved outward. the same happened for neon, carbon, helium and hydrogen. By the time iron starts forming at the core, you have multiple layers of fusion going on, one above the other. As the more and more silicon fuses, the iron core, ( where no fusion occurs) gets larger. This iron core is not supported by the heat of fusion, but by electron degeneracy pressure. If the iron core grow large enough, even this is not enough. Once electron degeneracy pressure gives out, the next step is neutronium. The new neutronium core is so much smaller than the iron core was that all the upper levels fall inward. The impact of these layers on the new core is what triggers the supernova explosion.

 

[Ken G]

OK, there are quite a few misconceptions in this thread that require correcting! The details get complicated of course so complete simulations are going to be better than these simple answers, but nevertheless there are some important basic concepts that are not coming out quite right. I'll try to provide some simplified corrections that are closer to the truth:

Core-collapse supernovae do not occur because the absence of fusion causes the core pressure to drop precipitously. In fact, the core pressure never drops precipitously in a supernova, it only finds itself substantially weaker than the hugely rising gravity. What happens is that whenever degenerate electrons go relativistic, they become very easy to contract, because the energy that is released by gravity during contraction is just what the electrons would need to keep them in their ground state as the core contracts to smaller radius. So the electrons become ambivalent to contraction-- the ground state's energy requirements are satisfied at any radius. This all depends on the mass reaching the Chandrasekhar mass-- that is the situation where degenerate electrons go relativistic.

So this means the pressure never drops, it simply doesn't increase as much as gravity does as the core contracts. The problem is that when the electrons are relativistic, anything that happens that allows gravity to get slightly larger than pressure will continue unabated, because the contraction that ensues will never release enough energy to get the pressure to rise enough to catch up to the rising gravity. Throw in energy loss mechanisms like processes that release neutrinos or that break up iron nuclei, and the rising pressure lags even more behind the much faster rising gravity.

The reason fusion prevents all this from happening is not that fusion is needed to keep the pressure high-- in fact fusion always keeps the pressure from rising higher because it prevents net heat loss and so prevents contraction. So what fusion actually does is prevent the net loss of heat that would otherwise produce the gravitational contraction that causes the kinetic energy to rise and the electrons to go relativistic. To have the standard type of core collapse, the electrons must always first be degenerate and relativistic, and both those things require net heat loss to occur, which fusion prevents. So the star must wait until the core is iron.

A couple other minor points-- type II supernovae are not the only core-collapse types, all the main types are core collapse except type Ia. Type Ia start out looking like a collapse due to degenerate relativistic electrons, but these stars contain fusable material, usually carbon and oxygen, and so when they contract they produce a sudden drastic amount of fusion, which completely unbinds the whole star and there is no neutron star or black hole created, it all blows out. Also, gamma rays do not heat up the matter in a core-collapse supernova, it is mostly neutrinos that do that. Neutrinos normally escape easily, but not in a core collapse-- that is the one situation where the density is so high it is capable of trapping neutrinos until they heat the gas. But even that is viewed as a kind of detail in the explosion-- most of the impulse comes from a "core bounce", which simply means that the gravitational energy that normally produces infalling motions gets reversed into outward motion. Remember, it is energy that is conserved in each radial wedge, not momentum-- momentum is a vector, and is zero over the whole explosion, so there's no issue with momentum conservation and core bounces are perfectly allowed if the core gets rigid enough as it neutronizes.

 

[D H]

OK, there are quite a few misconceptions in this thread that require correcting!

There are in turn quite a few misconceptions (and some key omissions) in your post that require correcting.

The key misconception is "This all depends on the mass reaching the Chandrasekhar mass-- that is the situation where degenerate electrons go relativistic." This is incorrect; electrons "go relativistic" well before the mass reaches the Chandrasekhar limit. The Chandrasekhar limit is the point at which electrons are ultra-relativistic (moving at the speed of light), not just relativistic. The key omission is that adding mass to a degenerate body makes that degenerate body shrink.There's no problem at first when a star first starts fusing silicon into iron. All that happens at first is that the iron that results from that fusion forms a degenerate core. A silicon shell around that degenerate iron core continues fusing, adding mass to the degenerate core. Adding mass to a non-relativistic degenerate object makes the object shrink slightly. If that's all that happened, the star would gradually build up a degenerate core of iron that gradually increases in mass and gradually decreases in radius.

That's not all that happens. Relativistic effects start kicking in when the mass of the degenerate iron core reaches 1/10 of the Chandrasekhar limit. These relativistic effects make the degenerate core shrink much more rapidly than a non-relativistic model would suggest. Increasing mass toward the Chandrasekhar limit means an electron's velocity approaches the speed of light and the radius of the body approaches zero.

The Chandrasekhar limit results from some simplifying assumptions. The derivation ignores the nuclei, ignores temperature, and ignores that the degenerate core might turn into something else (e.g., a neutron star) before the object shrinks to zero radius. As an approximation, it nonetheless is a very good one, and it shows what drives a core collapse supernova. The core collapses faster and faster as more and more mass is added to it, and this collapse becomes very rapid as the object's mass approaches the Chandrasekhar limit. This rapid collapse of the degenerate core is what drives the collapse of the rest of the star.

Except for the most massive of stars, the core collapse stops when the degenerate core changes from iron to something else (e.g., a neutron star). The rest of the collapsing star bounces off this neutronium core. This bounce is what makes the star "explode". As KenG mentioned, this bounce is not yet completely understood.

 

[Ken G]

There are in turn quite a few misconceptions (and some key omissions) in your post that require correcting.

I certainly would not call them misconceptions, merely different word usage. Methinks thou doth protest too much. Did I strike a nerve?

The key misconception is "This all depends on the mass reaching the Chandrasekhar mass-- that is the situation where degenerate electrons go relativistic." This is incorrect; electrons "go relativistic" well before the mass reaches the Chandrasekhar limit. The Chandrasekhar limit is the point at which electrons are ultra-relativistic (moving at the speed of light), not just relativistic. The key omission is that adding mass to a degenerate body makes that degenerate body shrink.

Well, perhaps you interpreted something in my remarks that I did not put in them. I did not intend any distinction between "ultra-relativistic" and "relativistic". The way to derive the Chandrasekhar limit is to simply insert the assumption that the ratio of pressure to kinetic energy density is the "relativistic" answer 1/3, rather than the "non-relativistic" answer 2/3. If you like, you can certainly parse those meanings some other way, maybe we could say the ratio is 1/3 when "ultra-relativistic" and somewhere between if only "relativistic". I had no real interest in such nuances of language, the Chandrasekhar limit is a conceptual approximation not a precise simulation, and the key point I was making is that no one else in the thread even mentioned that electron approach to c is the whole reason we have supernovae, and instead focussed on a non-existent precipitous drop in pressure after fusion ends-- that was the main correction I was offering. The rest of your remarks I agree with, but I don't see why you think they involve any "correcting", as they are completely consistent with everything I said.