Simulation Reveals Spiraling Supermassive Black Holes

A new model is bringing scientists a step closer to understanding the kinds of light signals produced when two supermassive black holes, which are millions to billions of times the mass of the Sun, spiral toward a collision. For the first time, a new computer simulation that fully incorporates the physical effects of Einstein's general theory of relativity shows that gas in such systems will glow predominantly in ultraviolet and X-ray light.

Just about every galaxy the size of our own Milky Way or larger contains a monster black hole at its center. Observations show galaxy mergers occur frequently in the universe, but so far no one has seen a merger of these giant black holes.

Scientists have detected merging stellar-mass black holes -- which range from around three to several dozen solar masses -- using the National Science Foundation's Laser Interferometer Gravitational-Wave Observatory (LIGO). Gravitational waves are space-time ripples traveling at the speed of light. They are created when massive orbiting objects like black holes and neutron stars spiral together and merge.

Supermassive mergers will be much more difficult to find than their stellar-mass cousins. One reason ground-based observatories can't detect gravitational waves from these events is because Earth itself is too noisy, shaking from seismic vibrations and gravitational changes from atmospheric disturbances. The detectors must be in space, like the Laser Interferometer Space Antenna (LISA) led by ESA (the European Space Agency) and planned for launch in the 2030s.

But supermassive binaries nearing collision may have one thing stellar-mass binaries lack -- a gas-rich environment. Scientists suspect the supernova explosion that creates a stellar black hole also blows away most of the surrounding gas. The black hole consumes what little remains so quickly there isn't much left to glow when the merger happens.

Supermassive binaries, on the other hand, result from galaxy mergers. Each supersized black hole brings along an entourage of gas and dust clouds, stars and planets. Scientists think a galaxy collision propels much of this material toward the central black holes, which consume it on a time scale similar to that needed for the binary to merge. As the black holes near, magnetic and gravitational forces heat the remaining gas, producing light astronomers should be able to see.

The new simulation shows three orbits of a pair of supermassive black holes only 40 orbits from merging. The models reveal the light emitted at this stage of the process may be dominated by UV light with some high-energy X-rays, similar to what's seen in any galaxy with a well-fed supermassive black hole.

Three regions of light-emitting gas glow as the black holes merge, all connected by streams of hot gas: a large ring encircling the entire system, called the circumbinary disk, and two smaller ones around each black hole, called mini disks. All these objects emit predominantly UV light. When gas flows into a mini disk at a high rate, the disk's UV light interacts with each black hole's corona, a region of high-energy subatomic particles above and below the disk. This interaction produces X-rays. When the accretion rate is lower, UV light dims relative to the X-rays.

Based on the simulation, the researchers expect X-rays emitted by a near-merger will be brighter and more variable than X-rays seen from single supermassive black holes. The pace of the changes links to both the orbital speed of gas located at the inner edge of the circumbinary disk as well as that of the merging black holes.

The simulation ran on the National Center for Supercomputing Applications' Blue Waters supercomputer at the University of Illinois at Urbana-Champaign. Modeling three orbits of the system took 46 days on 9,600 computing cores.

Credit: NASA's Goddard Space Flight Center

Music: "Games Show Sphere 01" from Killer Tracks

Follow NASA’s Goddard Space Flight Center

[NASAGoddard]

The blank circular region between the two black holes wasn’t modeled in this version of the simulation, which took the Blue Waters supercomputer 46 days on 9, 600 computing cores to produce. The reason it was not included was because the grid of points in space, or "3-d pixels", at which the physics equations are solved pinch together like thin pizza slices at the center of the region. Since the spacing between these points is so small and the dynamics are so fast, it requires the simulation to take very small steps in time to capture the motion across these small distances. This is called the Courant condition. Covering the center region means we would have to make the "pizza slices" even thinner and make our time steps even smaller, so our simulation would have to take many more animation steps to reach the same point in time. Future models will map the gas passing between the black holes in that region, but will require even more processing power or new methods of gridding (which are under development).

[Mr.Fox.]

The simulation ran on the National Center for Supercomputing Applications’ Blue Waters supercomputer at the University of Illinois at Urbana-Champaign.

Modeling three orbits of the system took 46 days on 9, 600 computing cores. 

Holy cow!!!

[Nico_Sama]

This would make for a cool animated desktop wallpaper if it wasn't for the text.

[alexpearson]

I’m always amazed at how natural these look.

[SuperYtc]

BREAKING NEWS: Simulation reveals what simulator simulated in said simulation.

[adamschultz]

Yet another example of how accurate simulations of the principles found in physics can prepare us and get us closer to finding anomalies like this.

[AndrejPodzimek]

I’m very curious about the region that wasn’t modeled. What’s in there? Was it prohibitively expensive to calculate? And, if it weren’t for the dangerous streams of super-hot gas, what would be seen by an observer flying through the barycenter between two (equally sized) black holes in very close orbit, but early enough before their event horizons touch? Would such a fly-by have all the “Interstellar style” time dilation effects? Would any other interesting and surprising phenomena occur?

[Javiermn]

That's crazy cool.




It's funny how many armchair physicists come out of the woodwork for videos like this and offer explanations that are as vague, obtuse and incomplete as they say the video is.

[shresthabageshwar]

This is the coolest thing I have seen in a while...

[OnaRocketship]

Question: I was watching another video like this and I noticed something. When this occurs is there a …opposite gravity force that splashed out? I noticed steep downhill means stronger gravity but I notice I positive bump or “hill” and I was curious to ask if that’s accurate observation.

[donnymcjonny]

Shoutout to whomever simulated this. Fantastic!

[NeilFiertel]

A marvellous and uncanny animation. It is a wonder that we now get to understand visually and from this continue to explore even from our own planet...

[strigoiimortii]

Gorgeouus, astounding, hypnotic. Pushing again button :)

[ganeshs]

I really wonder how these simulations are produced in a supercomputer...what is the STUFf on which the supercomputer wud produce those ecstatic results...?

[foutmaster]

If you look real closely, you can actually see Sans from Undertale

[Mr.Fox.]

So this shows 3 orbits that are 40 orbits shy of these two merging together, and it took 46 days to simulate this. (Read video description, very interesting)

[sacr]

So essentially, a Black Hole merger wouldn't necessarily release a supermassive "explosion" or release of thermal/radiation energy, it would just "boop" merge into one slightly larger black hole. Since the gravity on both ends is too intense to let anything go, a merger wouldn't release squat since everything can't escape that pull anyways. Right?

I'd go google it and find out, but i'm lazy atm.

[joeandcoffee]

This is amazing

Can you do simulation on what it looks like inside Black Hole ???

[dennisjames]

How do these black holes spin? Is it you making them spin ? Does the space time have some kind of elasticity?

[nikomaisuradze]

Great sim. Is space-time curvature considered during dynamics simulation or dynamics of accretion flow and light bending in space-time curvature are being modelled as separate processes and are later combined during rendering? Would really appreciate an answer, thanx in advance.