We were wrong about how Super Massive Blackhole get bigger!

[PhantomWolf]

Dr. Rebecca Smethurst, aka Dr. Becky on YouTube, has released her latest paper on her current research into Supermassive Black Holes and how they get bigger. Previously it was thought that most of their mass comes from the combining of the Black Holes when galaxies collide, but her team has discovered that this is incorrect and that the Black Holes in galaxies that have never collided are of similar size to those that have, meaning there was another mechanism. This paper discusses what they found and opens up some new theories about galaxy formation and evolution. If you are into astrophysics, have a look.

https://arxiv.org/pdf/2108.05361.pdf

 

[Puppycow]

I always assumed that they got bigger by swallowing up nearby stars, other black holes, neutron stars, and any other matter in the vicinity.

I'll go watch the video. I also wonder whether they formed in the Big Bang, or later. Did they begin as stellar-mass black holes that grew by swallowing up more matter, or did they somehow start out as supermassive black holes and only grow a little bit after that?

 

[PhantomWolf]

I always assumed that they got bigger by swallowing up nearby stars, other black holes, neutron stars, and any other matter in the vicinity.

I'll go watch the video. I also wonder whether they formed in the Big Bang, or later. Did they begin as stellar-mass black holes that grew by swallowing up more matter, or did they somehow start out as supermassive black holes and only grow a little bit after that?

I'm not sure that I can summarize the paper in a way that actually does it justice, but it seems that galaxies structures such as spiral arms and bars are systems that channel material and gas down into the centre of a galaxy, and it is this material that the SMBH feeds on, while also throwing a lot of it out again.

Dr. Becky talks about the paper and its finding in this video at 16:12...



More on Black holes with Dr Becky...

 

[rjh01]

If she is right then here is my prediction. If you look back in time the supermassive black holes would be much smaller than they are now.

 

[p0lka]

If she is right then here is my prediction. If you look back in time the supermassive black holes would be much smaller than they are now.

I predict they will get bigger in the future.

 

[lomiller]

I always assumed that they got bigger by swallowing up nearby stars, other black holes, neutron stars, and any other matter in the vicinity.

I'll go watch the video. I also wonder whether they formed in the Big Bang, or later. Did they begin as stellar-mass black holes that grew by swallowing up more matter, or did they somehow start out as supermassive black holes and only grow a little bit after that?

There is a limit to how fast a black hole can grow that way. The process of “eating” a star causes the a massive release of energy near the black hole as the start is torn apart, and this pushed nearby matter away from the black hole slowing down it’s growth. Apparently 13 billion years isn’t nearly long enough for supermassive black holes to get as big as they are just by eating stars and dust clouds.

 

[PhantomWolf]

There is a limit to how fast a black hole can grow that way. The process of “eating” a star causes the a massive release of energy near the black hole as the start is torn apart, and this pushed nearby matter away from the black hole slowing down it’s growth. Apparently 13 billion years isn’t nearly long enough for supermassive black holes to get as big as they are just by eating stars and dust clouds.

They need to be eating about an average of a sun's worth of material each year.

 

[Samson]

Let us assume when all the matter was in one place and not in far flung places, it was simpler for black holes to form.

 

[HansMustermann]

Let us assume when all the matter was in one place and not in far flung places, it was simpler for black holes to form.

Well, in fact simpler than to explain why the heck isn't the whole universe a black hole. There's still some debate as to whether basically the Schwarzschild radius for the mass of the visible universe is actually the radius of the visible universe. Now go far enough back in time when the same mass was packed in a tighter radius, and it definitely should have just gone black hole.

 

[Darat]

Off topic posts fed to our very own supermassive blackhole - AAH. It is estimated that within the next decade 99% of all posts will have disappeared into the black hole.

Replying to this modbox in thread will be off topic  Posted By: Darat

 

[Mike Helland]

There is a limit to how fast a black hole can grow that way. The process of “eating” a star causes the a massive release of energy near the black hole as the start is torn apart, and this pushed nearby matter away from the black hole slowing down it’s growth. Apparently 13 billion years isn’t nearly long enough for supermassive black holes to get as big as they are just by eating stars and dust clouds.

How long is long enough?

 

[Ziggurat]

Well, in fact simpler than to explain why the heck isn't the whole universe a black hole. There's still some debate as to whether basically the Schwarzschild radius for the mass of the visible universe is actually the radius of the visible universe. Now go far enough back in time when the same mass was packed in a tighter radius, and it definitely should have just gone black hole.

The larger a black hole is, the less dense it is, if you calculate the density by dividing the mass by the volume inside the Schwarzchild radius. So at any given density, you can calculate a radius that would give you a black hole. If the universe has some particular density at some particular time, you can calculate a corresponding Schwarzschild radius, and see if that radius is smaller than the universe is. And if you take this approach, it may indeed look like the universe should have turned into a black hole.

But this approach is wrong.

The thing about the Schwarzschild radius is that it's part of the Schwarzschild solution, and the Schwarzschild solution is for a spherically symmetric mass distribution in an asymptotically flat universe. And this solution works pretty well for black holes we know of because at the scale of individual stars and even galaxies, the universe is reasonably close to flat.

But the universe is not actually asymptotically flat, especially the early universe. The fact that the universe is expanding makes the Schwarzschild solution completely invalid at the kind of scales you're describing. So you cannot use the Schwarzschild solution when evaluating the universe as a whole. It doesn't work.

 

[HansMustermann]

Well, the thing is, if you go back in time far enough, the whole mass was concentrated in something the size of a galaxy or indeed of a (really big) star. So if calculating the Schwarzschild radius works at that scale, we still have a problem: the mass in the universe at that point is in fact creating a Schwarzschild radius larger than the actual radius of the universe. Again, at that point in time.

"But that radius is very big, compared to the size in which the mass is" doesn't really work, because it doesn't work for the singularity in a black hole.

It seems to me like Birkhoff pretty explicitly says that the solution of the gravity of that sphere WILL be asymptotically flat, and that whatever growth or shrinking happens in the size of that sphere, is not going to matter. Well, at least as long as it's reasonably spherical, non-rotating, etc.

So SOMETHING must have still pushed hard outwards for it to grow instead of just implode right back.

 

[Ziggurat]

Well, the thing is, if you go back in time far enough, the whole mass was concentrated in something the size of a galaxy or indeed of a (really big) star. So if calculating the Schwarzschild radius works at that scale, we still have a problem: the mass in the universe at that point is in fact creating a Schwarzschild radius larger than the actual radius of the universe. Again, at that point in time.

No, that's not how it works. Right now, space is close enough to asymptotically flat for the approximation to work for galaxies. But in the early universe, when huge amounts of matter were within the volume of a galaxy, the curvature of the universe was also larger. On the scale of a galaxy-sized object, the universe was not asymptotically flat. The scale at which the flatness approximation breaks down isn't fixed. It's smaller the farther back we go. So yes, everything was denser in the early universe, but everything was also more curved.

It seems to me like Birkhoff pretty explicitly says that the solution of the gravity of that sphere WILL be asymptotically flat, and that whatever growth or shrinking happens in the size of that sphere, is not going to matter. Well, at least as long as it's reasonably spherical, non-rotating, etc.

No. Asymptotic flatness isn't a result of the solution, it's a required input. Without that input, you cannot get that solution.

 

[HansMustermann]

No, that's not how it works. Right now, space is close enough to asymptotically flat for the approximation to work for galaxies. But in the early universe, when huge amounts of matter were within the volume of a galaxy, the curvature of the universe was also larger. On the scale of a galaxy-sized object, the universe was not asymptotically flat. The scale at which the flatness approximation breaks down isn't fixed. It's smaller the farther back we go. So yes, everything was denser in the early universe, but everything was also more curved.

Isn't that how it works for a black hole, though? If it's curved enough, you get an event horizon. I'm not sure how something can be too curved to be a black hole, if we're talking mass alone. (But of course it can be if some other effect curves it the other way around.)

No. Asymptotic flatness isn't a result of the solution, it's a required input. Without that input, you cannot get that solution.

It may be my misunderstanding of Birkhoff, but it seems to me like nowhere does it require asymptotic anything inside the sphere. The whole point is that the space-time distortion of any spherical and non-rotating distribution of mass can be calculated as being a point mass in the centre. It doesn't matter if the distribution by radius is uniform, asymptotic from the centre, all below surface of the sphere, or whatever.

Also, exactly what does Birkhoff say, if asymptotic flatness
1. outside the sphere,
2. of the effect of the mass inside the sphere,
apparently isn't it?

 

[Ziggurat]

Isn't that how it works for a black hole, though? If it's curved enough, you get an event horizon. I'm not sure how something can be too curved to be a black hole, if we're talking mass alone.

The Schwarzschild solution is a solution for a black hole sitting in a flat spacetime. The spacetime far away from the black hole must be flat, the spacetime near the black hole is obviously not.

If the spacetime far away from the black hole is NOT flat, then the Schwarzschild solution is not valid.

It may be my misunderstanding of Birkhoff, but it seems to me like nowhere does it require asymptotic anything inside the sphere.

I'm not talking about flatness inside the sphere. Asymptotic flatness means that spacetime approaches perfect flatness as you go out infinitely far away. That's the whole point. The exterior solution is what must be asymptotically flat. But the universe isn't.

I think what may be going on is that you're misunderstanding the Wikipedia entry, or some other source that's copied from it. So let's take a look at it for a moment.

In general relativity, Birkhoff's theorem states that any spherically symmetric solution of the vacuum field equations must be static and asymptotically flat.​

This makes it sound like asymptotic flatness is an output of the theory, not an input. But that's not quite true. And the reason it's not really true is hiding in this very sentence, though it's easy to miss: the vacuum field equations. What does vacuum mean here? It means no mass.

So what we're looking at with this theorem is a spherically symmetric solution in which there's no mass in our region of interest. There can be mass somewhere, but (1) that mass must be spherically symmetric so our solution is as well, and (2) we're only looking at the solution outside of that mass, and there is no mass in this exterior region.

But that lack of mass anywhere else is what leads to asymptotic flatness. By saying that there's basically no mass anywhere except in this specific region, we've basically made asymptotic flatness an input, just under a different name.

And when we're talking about cosmology and what's happening at very large scales, you're obviously NOT talking about the universe having no mass outside of a particular region. There's mass everywhere. You can't get away from it. So even ignoring asymptotic flatness as such and using the phrasing as presented by Wikipedia, Birkoff's theorem is invalid from the start for examining cosmology because we can't use a vacuum solution when the universe isn't a vacuum.

 

[HansMustermann]

Still not sure how that's supposed to work. Whether it's GR or even newtonian mechanics, sure, mass outside can add its own distortion. Like, we can calculate the gravity of the Sun at 150 million km to a value, and yes, it's calculated as if there was a void outside the Sun, but if you're on Earth chances are it's not the dominant component. Doesn't prevent us from calculating the gravity of the Sun at that distance that way anyway. In fact, at least for classic gravity, it's kind of the whole point of the shell theorem.

 

[HansMustermann]

That said, it seems to me like an easier case could be made for it just being an effect of the universe being actually infinite or a closed manifold, rather than trying to make Birkhoff not apply, innit? But that goes through making the space flatter in every point (at least as far as mass contribution to that curvature goes), rather than too curved.

 

[Ziggurat]

Still not sure how that's supposed to work. Whether it's GR or even newtonian mechanics, sure, mass outside can add its own distortion. Like, we can calculate the gravity of the Sun at 150 million km to a value, and yes, it's calculated as if there was a void outside the Sun, but if you're on Earth chances are it's not the dominant component. Doesn't prevent us from calculating the gravity of the Sun at that distance that way anyway. In fact, at least for classic gravity, it's kind of the whole point of the shell theorem.

Well, let's take a look at the classical Newtonian gravity case for a moment.

Let's suppose we're in a Newtonian universe, where space is infinite and mass is distributed uniformly throughout that universe.

Now let's take a sphere of radius R, with us positioned on the right side of that sphere (so the sphere is to our left). The mass within that sphere will pull on us, and so we should feel a gravitational pull to the left from that sphere. OK, but what about the mass from the rest of the universe? Well, for that, we can break the rest of the universe into concentric shells of mass. And we're inside each of those shells. So we should feel no gravitational attraction. That means we should feel a net force to the left.

But what if, instead of starting with a sphere on to our left, we start with a sphere to our right? We go through the same process, and determine that we should feel a net force to the right.

There's a conflict here. How can we resolve it? What the hell is going on?

What's going on is that the shell theorem is essentially giving us an infinite series, and the terms of the infinite series aren't actually getting smaller. The series can appear to converge on different answers, depending on how you order your terms. This is a strong indicator that that's not the right approach.

The same will be true in general relativity. Each unit of farther away mass will have a smaller effect, but there's also more mass farther away. In Newtonian mechanics, the terms of any one shell will always cancel, and you're OK as long as you don't go infinite. But Newtonian gravity is a linear theory. General relativity is not linear. And while general relativity does still have its equivalent of a shell theorem, it isn't the same shell theorem. Things don't simply cancel.

If you are working on a mass density scale where the universe as a whole has a similar density to the area you're interested in, you cannot ignore the gravitational effects of the rest of the universe. We can ignore the gravitational effects of the universe as a whole when looking at, say, the orbits of the planets in our solar system, because the mass density of the solar system is so much higher than the mass density of the universe as a whole. But that obviously won't work when you're talking about why the observable universe itself isn't a black hole, because the universe beyond the observable universe is going to have similar mass density to the observable universe.

 

[Mike Helland]

Well, let's take a look at the classical Newtonian gravity case for a moment.

Let's suppose we're in a Newtonian universe, where space is infinite and mass is distributed uniformly throughout that universe.

Now let's take a sphere of radius R, with us positioned on the right side of that sphere (so the sphere is to our left). The mass within that sphere will pull on us, and so we should feel a gravitational pull to the left from that sphere. OK, but what about the mass from the rest of the universe? Well, for that, we can break the rest of the universe into concentric shells of mass. And we're inside each of those shells. So we should feel no gravitational attraction. That means we should feel a net force to the left.

But what if, instead of starting with a sphere on to our left, we start with a sphere to our right? We go through the same process, and determine that we should feel a net force to the right.

There's a conflict here. How can we resolve it? What the hell is going on?

What's going on is that the shell theorem is essentially giving us an infinite series, and the terms of the infinite series aren't actually getting smaller. The series can appear to converge on different answers, depending on how you order your terms. This is a strong indicator that that's not the right approach.

The same will be true in general relativity. Each unit of farther away mass will have a smaller effect, but there's also more mass farther away. In Newtonian mechanics, the terms of any one shell will always cancel, and you're OK as long as you don't go infinite. But Newtonian gravity is a linear theory. General relativity is not linear. And while general relativity does still have its equivalent of a shell theorem, it isn't the same shell theorem. Things don't simply cancel.

If you are working on a mass density scale where the universe as a whole has a similar density to the area you're interested in, you cannot ignore the gravitational effects of the rest of the universe. We can ignore the gravitational effects of the universe as a whole when looking at, say, the orbits of the planets in our solar system, because the mass density of the solar system is so much higher than the mass density of the universe as a whole. But that obviously won't work when you're talking about why the observable universe itself isn't a black hole, because the universe beyond the observable universe is going to have similar mass density to the observable universe.

At least in classical mechanics does it make sense to put a shell around the entire universe?

That would imply you're calculating the effects of gravity on something not in the universe. So... pretty dead end, aint it?