This has curious similarity with errors which can appear in basic n-body simulations, where excessively close interactions can break orbits by throwing bodies too far from circumstances in one timestep to the next.
In n-body simulation the 'point gravity' objects are kind of mathematical singularities which never collide because they have no size. The integration time-step gives them a kind of radius or gradient of proximity where very large errors occur, where virtual energy can be added or lost.
I read in these cases of true black hole collision, the shedding of energy through gravitational waves is the main cause of their orbits decay - without making the waves, the holes could circle each other for a very long time. Presumably energy&momentum is maintained over the collision of the two astrophysical non-point singularities, unlike the basic n-body excessive gravitation phenomenon.
The full details of what actually happens during the collision/overlap of black-hole 'calculation radii' are surely mysterious.
this is also my experience with simulated gravity and galaxy systems. i guess newton physics alone would be enough to make a body flung out if it comes close enough. and i guess a merge is close enough.
i was working on a space mmo and after leaving the server running for weeks some body would always be flyng away with crazy speed.
The newtonian model itself doesn't produce this behaviour as it is continuous, so however much bodies accelerate towards each other they can decelerate just as much once they have passed. In n-body simulation the continuous model is 'squeezed' into time-slices, where bodies can accelerate towards each other in one time-slice and in the next be too far past each other to slow each other down again.
Seemingly in the real astrophysical world, blackholes will
free-fall towards each other by their masses 'dragging' on spacetime, causing deep perturbations in spacetime (as gravity waves). Eventually there is a 'merge' as you called it, an operation during which the resulting merged body can get a directed kick from the persisting spacetime perturbations, one powerful enough to eject it from a galaxy.
In the real world, overall it is expected energy and momentum is not defied during this merge operation, it seems remarkable considering how exceptional and energetic these events appear.
edit
...on second thoughts, this 'merge event' produces GRAVITATIONAL PROPULSION - sorry for the caps but I think they might be justified.
In Newtonian gravity you need at least three bodies to arrange an ejection: two of them start orbiting more tightly, to counterbalance the third one flying out to infinity. (IIRC this happens a lot in globular clusters, and simulations like yours.) With just two bodies you get only Newton's solutions, the conic sections.
This make me wonder if the galactic void is full of ejected black holes which will be nearly undetectable since the X-ray radiation comes from the accretion disk and the matter falling into the black hole.
This mechanism could be potentially be another candidate for "dark matter".
This is a good observation, but gravitational lensing surveys have all but eliminated massive compact objects as a potential source of dark matter; see for example https://www.ncbi.nlm.nih.gov/pubmed/17359015.
The NIH (specifically PubMed) just indexes most scientific journals. This paper is actually in Phys. Rev. Letters (which the NIH site links to, if you click the DOI link below the abstract):
Gravitational lensing surveys coverage is pretty darn small so far it's also might not be appearant on the scales we are talking about there will need to be a pretty lucky alignment of a galaxy and a black hole given the space and relative size of both objects and the noticeable lense diameter I'm not sure that intergalactic lensing will be detectable as intragalactic ones which also are pretty rare.
Not undetectable. There is still gravity lensing. And black holes (the small ones) are still stars, or at least mass/matter that once was. That material still has a place in the equations for the evolution of stars in galaxies.
Oops, yeah that was a goof... for some reason my brain interpreted that as dark energy. I even wrote "dark matter" and didn't catch myself. Forget I even said anything ;)
Amazing - the calculations in the article suggest gravitational waves emitted by a pair of colliding black holes can have sufficient energy to eject the combined object from the potential well of the galaxy.
Actual headline has "... out of galactic core". The black hole is still in the galaxy, though perhaps in 20 million years it will have escaped.
The involvement of gravitational waves is conjecture and the evidence for it is highly indirect (it all sounds pretty plausible to me, but what do I know?).
Gravitational waves are among the most difficult physical phenomena to detect directly, because of the very small size of the effects involved and the huge masses required to cause those small effects (one commonly encounters numbers like 10^-51 when doing gravitational wave calculations). The original experimental demonstration of gravitational waves was from analyzing the change in rotational frequency over time of an orbiting binary pulsar and computing that the energy loss in the system was precisely matched by the general relativistic prediction of energy radiated away in gravitational waves. The importance of this work was recognized with the Nobel Prize in Physics [0].
Can someone please explain why such waves that supposedly have minimal/hard-to-detect effects are somehow valid explanations for pushing around black holes?
Meanwhile, electrical forces are clearly powerful and yet still neglected in the mainstream.
1. Electromagnetic radiation has been the primary tool of astronomers and astrophysicist since before the first telescope. It is not "neglected" by any stretch of the imagination.
2. At large scales, the total number of electrons and protons tend to balance and thus result in electrically neutral objects.
Black holes by definition can't emit EM radiation, so their only way to emit energy (ignoring hawking radiation which is tiny) is through gravitational waves. Due to the conservation of momentum, to be ejected from the core must be a result of some amount of momentum in the opposite direction.
(Napkin math time.) A black hole with 1 billion solar masses would pull Earth with a force equal to the Sun from half a light year away. It does the same thing to Uranus from a distance of 9.6 light years. It would have enough influence to disturb those orbits from much further away.
So if something similar entered our galactic neighbourhood, we would notice changes in the proper motion of stars closer to it, discrepancies between the planets' predicted and actual positions, and (possibly before anything else) predictable errors in GPS timing (since this is probably the most commonly-encountered system depending on accurate knowledge of speed, time and gravity).
What I'm wondering is, if this is the explanation, why don't we see this more often? Plenty of galaxies have formed from mergers, right? So the question is, how much of a difference between the black holes does there need to be for this to occur? The video seems to imply only a small one, in which case one might expect this to be a common occurrence. So why haven't we seen this before? Or is that not the case?
Because we would need to be looking. The sky is so very vast, and our imaging equipment needs to look at such a small part in order to resolve the galaxies beyond our own.
Not much. Relative to a star, central supermassive black holes are, well, super-massive. But relative to the total mass of the galaxy, or even just the core region, they aren't that big. The rest of the galaxy will just keep orbiting around its center of gravity, regardless of whether or not there happens to be a black hole there.
Pretty much any galaxy. The Milky Way has something like 300 billion stars, with the "average" star weighing 1/4 as much as the sun. So even the stellar mass of a galaxy like ours outweighs such a black hole by two orders of magnitude, and most of a galaxy's mass is dark matter.
Well, for reference, the lower bound on estimates of the mass of the milky way is ~5×10^11 solar masses, so a billion solar mass black whole would account for less than half of a percent of that.
Truthfully, most if not all large galaxies probably do have central supermassive black holes. Dwarf galaxies might not, but they're more like drifting clouds (and tend to be gobbled up by larger galaxies over time). It's just that there is no way to say with certainty that a BH is in the center of a galaxy when it isn't actively feeding, unless it's close enough for us to resolve the motions of individual stars in the core.
Remember also that there are large swaths of the sky we basically cannot observe in the 'Zone of Avoidance'.
If a galaxy is in a position to be lensing, or we can otherwise get a grip on its mass, then we can guess the BH if it isnt emitting enough. If the galaxy is dim but massive, black holes are likely.
My guess is that it gains a new black hole as stuff coalesces at the center. This hunch comes from not much more than knowing what Newton predicted, which is that any system with more than two bodies is unstable, with outcomes that include ejecting a body, or two bodies colliding. The formation of stars in the first place is the result of such coalescence, so it's not unthinkable that the same thing will happen at the center of a galaxy full of stars.
What is the probability that the ejected black hole (if it is indeed such a thing) goes on to spawn a new galaxy?
Black holes are likely responsible for galaxy formation and evolution [1]. Since new black holes don't spin up from the void, I had assumed this meant all the galaxies to ever be formed have been formed (excluding effects of mergers). This, however, produces a new origin method (even if extraordinarily low-probability).
By warping spacetime. It's the same thing that causes us to fall toward Earth. The presence of Earth's mass warps spacetime, and thus the Earth imparts momentum on us.
Anywhere I can look for a more detailed explanation of the process by which a gravitational wave imparts momentum, though? It seems like the mechanism must be quite different from the geodesics in a constant gravitational field.
The nasa.gov link summarizes the speculation about a large anisotropic burst of gravitational radiation from the merging smaller precursors to the runaway supermassive black hole. One is free in GR to consider this as uniformly radiated GWs carrying away residual momentum along the x^1 and x^2 spacelike axes and most of the momentum from the spacelike x^3 axis as the merging black holes "bald" their asymmetries in line with the no-hair hypothesis. (The ^1..^3 are indices in the Einstein convention).
We can generalize somewhat, to answer your more broadly worded question.
First let's start with a system losing energy-momentum via gravitational radiation (e.g. PSR1913+16, but isolated in vacuum asymptotically flat spacetime) and under time-reversal[1], so that the system is gaining energy-momentum from gravitational radiation arriving from infinity in such a way that the orbit outspirals. (This requires a highly improbable spacetime towards (time-reversed) past infinity).
In this time-reversed picture, if you alter the incoming gravitational radiation, the energy-momentum imparted to the binary system will necessarily perturb their orbit differently; instead of causing the line segment through the barycentre between the pulsars' centres-of-mass to grow along its length with carefully timed + linearly polarized GWs[2], you can cause the entire system to move in a particular direction.
Think of incoming gravitational radiation that is highly anisotropic but arriving with lucky enough timing that the GW (in a gauge in which it is a plane wave with x linear polarization seen from "above" or "below" the plane of the binary orbit[2]) is always extended along this axis by stretch-squashing the advancing and receding bodies differently. An inertial observer in this asymptotically flat spacetime at a great distance and not feeling the GWs will see the system precess.[3]
Chapter 7 (Perturbation theory) in Carroll's _Spacetime and Geometry_ in subchapters 7.4-7.6 has a reasonably good inductive view of the mathematics in the limit of weak gravity, and the time-reversal trick should be workable with the contents of the chapter.
You could introduce anisotropy in a physically plausible (barely; Hulse-Taylor's orbital period is just a few hours) way by putting the binary pulsars near a pair of carefully arranged inspiralling supermassive black holes.
Another approach would be to consider Gravitational Bremsstrahlung ("GB") (see e.g. (d) at Kovacs & Thorne [1978] at http://articles.adsabs.harvard.edu/cgi-bin/nph-iarticle_quer... which has an extremely priceless admission in its final sentence.). The tl;dr of "GB" is that just as electromagnetic bremsstrahlung changes the momentum and direction of a small charge moving past a large charge with the emission of electromagnetic radiation, a small mass moving past a large mass will emit gravitional waves. One can again time-reverse in each case.
The usual case of electrons decelerating in a medium and throwing off photons under time reversal is electrons absorbing photons and accelerating in a medium. Likewise (given a suitable frame of reference and gauge, and in analogy with cyclotron radiation as a form of electromagnetic bremsstrahlung) a small mass body deflecting around a much larger mass decelerates and throws off gravitons, but under time reversal absorbs gravitons and accelerates instead.
Finally, choosing particular (families of) observers and considering things in a frame of reference constructed on their attributes, choosing different sets of coordinates and doing gauge-fixing is normal in GR; some people hate it because relating one set of observations under particular choices like these to the observations of the same system under different choices is either hard to intuit or conversely hard to solve the equations for (and sometimes both!). I hope the above isn't a total confusing mess.
[2] this is seriously (overly) simplified :) in the ordinary non-time-reversed picture the orbiting system is always shedding off a spectrum of gravitational waves and the plane of polarization changes twice per orbit; I instead focus on the instantaneous state where an observer above the plane of rotation only looks twice per orbit, seeing the system with one star at 12 o'clock and the other at 6 o'clock, and
[3] the whole "clock" precesses against the distant observer's cartesian coordinates with an origin constantly on herself.
P.S: there are some more article-relevant gory details at http://iopscience.iop.org/article/10.1086/421552 ("How Black Holes Get Their Kicks"). Note the authors raise the analogy with bremsstrahlung (where the masses are very different) at the bottom of page L6 - page L7.
It was very generous of you to give such a detailed explanation. Thank you.
> I hope the above isn't a total confusing mess.
Not at all. Pretty confident that I understand the explanation in [3], though not sure how that applies to a single mass but you've given me plenty of places to look for an explanation.
If the black hole got propulsion from this wave, I'm pretty sure it's not hard at all. You just have to be in the vicinity - no special shape or structure required.
In n-body simulation the 'point gravity' objects are kind of mathematical singularities which never collide because they have no size. The integration time-step gives them a kind of radius or gradient of proximity where very large errors occur, where virtual energy can be added or lost.
I read in these cases of true black hole collision, the shedding of energy through gravitational waves is the main cause of their orbits decay - without making the waves, the holes could circle each other for a very long time. Presumably energy&momentum is maintained over the collision of the two astrophysical non-point singularities, unlike the basic n-body excessive gravitation phenomenon.
The full details of what actually happens during the collision/overlap of black-hole 'calculation radii' are surely mysterious.