Two 50 Galaxy Cluster Crash Simulation (beta=90)

[PelicanGodOfJupiter]

@rberring Does this simulation take into account the effect of dark matter, or is it purely based on gravity?

[typesix]

It must take some hard core processing power to run a simulation like this. Badass!

[rberring]

@typesix By todays standards, it requires relatively little computational power. There are simulations that are many times larger like the millennium simulation that are much larger. Even the millennium simulation can be run in a relatively short time now. These simulations are within the capability of most desktop computers with required computational times ~a few days. Glad you enjoyed the simulations.

[CrazyPilotEarl]

The end result is what I imagine the "Great Attractor" to be.

[rberring]

@CrazyPilotEarl Well the current idea of what the great attractor is thought to be is a missing cluster of rather large mass ~10^{15-18} solar masses that has perturbed the flow of matter in the nearby universe. Some suggestions have been offered for possible sources, but as far as I know still eludes a definitive identification. But in some sense you are correct. However the "missing" cluster is probably much larger than the resulting cluster above.

[Volgeirea]

mr t vs chuck norris



[rberring]

@Volgeirea I'm not sure what to what you are referring or what relation this has to the simulation on this page.

[Volgeirea]

@rberring its a joke, relating what would happen if mike tyson and chuck norris got into a fight, even though this does not compare in the least bit, its close enough.

[mastanugget]

if 2 star clusters collided it would fling galaxies all over the place.

[rberring]

@mastanugget Well, it is first important to appreciate the scale in these simulations. A star cluster is a small component (a subset) of a galaxy, and you are seeing a cluster of galaxies. Each galaxy will have many star clusters within them. Now if two clusters of galaxies collided, the chances are good that it would send galaxies flying into the local surroundings, but that depends on the energetics of the collision, and you can easily see this in the simulations.

[mastanugget]

@rberring Gravity will be so great at the collision point that it might cause major star birth from the dust being torn from galaxies. It would be a massive train wreak in the sky.

[rberring]

@mastanugget Galaxy collisions have been known to cause star formation in galaxies. There are many examples of galaxy-on-galaxy collisions where massive star formation is occurring. The most likely culprit is the collision between the galaxies. So in short, a collision between two galaxies is effectively a massive train wreck in the sky.

[plsspareme1]

this is scary. imagine urself there watching another galaxy close in. .....

[rberring]

Actually you probably wouldn't notice much. The sky wouldn't look much different than it does today. But not to alarm you, the Andromeda Galaxy--the other major galaxy in our local group is headed straight for for us. But don't worry, the collision between the Milky way Galaxy and the Andromeda Galaxy won't happen for another ~100 million years or so.

[shutup735]

@rberring and besides, the odds of our solar system ending from that are only about 13-1, stars a far apart in galaxies...

[rberring]

In fact 13-1 is being VERY generous. It is more like 1,000,000,000,000-1. It is extremely unlikely that anything will happen to the solar system. What I find cool is if you were to come back in ~50,000,000 years the Andromeda galaxy would be much larger and brighter on the sky. It would be a spectacular sight.

[shutup735]

@rberring well maybe but if a star even comes within 1 trillion KM or 0.1 lightyears it could screw up the outer planets and cause millions of asteroids to random whip thru the solar system, earth would be pelted with asteroids and there will be very little survivers, but i will let my great great great great great great great great great great great, Ect grand children deal with that

[DillynWinn]

this isn't accurate, the clusters wouldn't pull towards eachother, and the galaxys wouldn't be flying around like that, clusters are just groups, they dont have their own gravity.

[DillynWinn]

I know there is a lot of gravity in a cluster, but if the cluster had it's own gravity strong enough to pull others towards it, then the galaxys in that cluster would be pulled towards eachother too and then they would all crash into eachother and die before there was enough time to pull another cluster towards it. I can see that two or more clusters could crash into eachother by accident, but it wouldn't have anything to do with gravity, just bad luck.

[rberring]

I'm not sure what you mean by die, but galaxies are pulled toward each other, and do interact with eachother. Over the age of the universe, on average a galaxy in the cluster environment will see ~3 close interactions with another galaxy in the cluster. While galaxy and galaxy cluster evolution is stochastic, it is gravity which dominates and drives the interactions. What you are seeing is the evolution of a cluster driven by gravity. In that sense, the simulation is accurate.

[DillynWinn]

I mean destroyed, and I can understand that they can pull towards eachother, but they would pull towards themselves too if they had the power to pull from that far and then they would just condense and crash into itself. that's like hanging a bunch of magnets really close together, withing inches, and then 5 feet away doing the same thing, and expecting the magnets to pull on the other group before it pulls into itself, understand what I mean?

[DillynWinn]

it's entierly possible that it would look similar to that if two clusters did crash into eachother, but the galaxies wouldn't be shooting around like that. and at the end it looked like that the two clusters were just galaxies themselves. and that all the galaxies were just planets.

[DillynWinn]

there weren't even any explosions from the galaxies coliding.

[rberring]

Ok, it is obvious there are many misconceptions at play here. It is pretty clear you are not understanding what you the simulations are. First there is no gas in these simulations. It is only gas and dark matter (see description). As a result no explosions will be produced. Secondly, what you are sayinga bove is not making sense. This is a simulation of a cluster of galaxies. Each white grouping is an entire galaxy. Planets are not not resolved in this simulation.

[rberring]

I am not sure what you are meaning by "shooting around like that." The galaxies are following their orbits in the gravitational potential well described by the mass distribution in the cluster. You are making an keen observation that the cluster look like at the end the cluster look like large galaxies. That is the natural evolution of a cluster it to result in a large centrally located galaxy at the center of the cluster.

[rberring]

In regards to this statement it is important to understand that electromagnetic fields while they follow inverse-square law have negative/positive (north/south) polarities. There is no negative mass, and thus magentic fields are not a good analgy to describe the behavior of gravitational laws. Despite what you are saying makes little sense.

[DillynWinn]

let's just stop arguing, I'm tired of trying to explain it to you, you don't understand at all.

[rberring]

Well, I would suggest taking an Astronomy course if you wish to learn more about the dynamics of galaxies clusters. A youtube page is not the forum to learn such a complicated subject.

[DillynWinn]

Yeah, I love Astronomy, but I can agree that this isn't the proper place to argue about it.

[rberring]

I am not sure how this is relevent with the simulation, but I see no indication why the Sun/solar system would not be born of the Milky Way.

[ToadHypno]

Our sun may not be indigenous to the milkyway galaxy...

[rberring]

I am not going to support or deny, but I will say I do find it hard to believe that the Sun is not indigenous to the Milkyway Galaxy. The fact that it is so close to the galacitc plane with a highly circular orbit tells me that the chances of the Sun being indigenous to the Galaxy is high.

[mallamoozoo]

it great fun to get stoned and watch stuff like this - I can't get enough! It's so ....INFINIE!

[rberring]

It is actually funny that you say that becasue it is. However, I thought mitosis was a process by which cells divided. These clusters are merging. So it is like watching cellular mitosis in reverse.

[unurautare]

what happens to the galaxies when the centers of the clusters try to merge? its like a galaxy colision and some galaxies explode?(lol) :P

[rberring]

If you mean my being ejected from the system when you say explode, you are correct. Much of the cluster orbital energy goes into ejecting galaxies from the system. The remainder of the energy goes into a restructuring of the internal dynamics of the cluster.

[sim5191]

if it's about the universe, i'm interested.

[sim5191]

i have to tell ya. if there's new info about the universe, I WANNA KNOW I'T!

[rberring]

We are always figuring out new things about the universe everyday. It is always exciting to meet people who share a common interest. Hope you enjoyed the movie. I enjoyed making it. Be sure to keep your interest in this fascinating subject.

[sim5191]

but ur right. some galaxies become new ones and some are thrown in the emptyness.

[sim5191]

fastest reaction ever had on youtube in my life!

[rberring]

Don't worry. I don't always respond this quickly. And thanks for your question. It was a very astute observation, and an excellent question.

[sim5191]

new cluster isn't 100 galaxies if two of 50 collide. some galaxies are "thrown away"!

[rberring]

You are correct. Some of the cluster orbital energy goes into the orbital energy of the galaxies in the cluster. Some of these galaxies do escape from the system. And also don't forget mergers occur. So the final cluster has fewer than 100 galaxies despite it being formed from two 50-galaxy clusters.

[Selassei69]

damn it! speak in spanish.

[rberring]

This is in reference to?

[Selassei69]

comments.. mostly ppl speak spanish so dont understand

[looseskrewz]

Is your code available?

and....thanks for the reply.

[rberring]

Actually the code is not mine. It was written by Lars Hernquist (and possibly Joshua Barnes). The code is freely available. However, it is written in Fortran 77, and if you don't have a Fortran compiler, you will either have to convert it to a language that you do have a compiler for, or you will not be able to use it.

[looseskrewz]

What are you guys using to simulate this?

I just got the CUDA nbody for my NV8800. It seems very similar, although there are a few features I would like to add, one would be being able to save a simulation.

If you can give away the secret please shoot me a messege or reply.

Thanks

[rberring]

The code used to simulate the above simulation is a Hernquist Treecode, and uses a spatial sorting technique to approximate gravitational forces. As I understand it, the CUDA Nbody is a direct nbody code. The above simulation has 160,000 particles and would be difficult to simulate in a reasonable amount of time on a desktop machine which is what is used for the simulation.

[rberring]

I have only recently heard about the CUDA code. I am guessing that it is possible to implement a treecode into the CUDA program. I think the idea of using GPUs for Nbody calculations is fairly recent.

[nxmoparmatt84]

isnt the distance between the stars great enough to where you could say its highly unlikely there will be any collision between stars or bodys in the galaxys initial crash?

[rberring]

Sorry for not getting back to you sooner, but you are correct. The chances of a stellar collision occurring as a result of a galaxy merger event is tiny. Despite the typical number of stars found in an average galaxy ~10^11 stars.

[nxmoparmatt84]

no prob

[maxvondel]

Yes the stars are far apart but the dust and gas are not. This in turn affects the stars embedded in them.

[rberring]

The gas will collide in a galaxy merger. In fact a merger event is thought to trigger star formation by the colliding gas clouds. The stars act like bullets that penetrate the gas clouds and as for the galaxy dynamics has little effect for one major reason: the dynamics are dominated by the distribution of dark matter.

[AndreCbrito]

this could be a process of how may years?

[rberring]

The simulation is for a Hubble time or the age of the universe. So for the currently fashionable model that is about 13 billion years (13,000,000,000 years).

[kwakhed23]

are the clusters virialized before the start of the simulation? it appears that you let it run until they have become so after the collision...

[rberring]

Yes, the cluster as a whole and individual galaxies are virialized at the start. I also evolve each galaxy in isolation to check for instabilities from say two-body relaxation. At the end of the simultaion it is difficult to say if the cluster as a whole is virialized. I am sure it is out to a radial distance (which is unknown to me at the moment), but beyond I suspect not.

[spookler]

Those are some odd looking galaxies... resembling something from my screensaver lol.

I recall viewing a documentary on this, apparently we're either going to be ejected into cold space or be swallowed by the supermassive blackhole. Either way we're screwed. :)

[rberring]

All the galaxies in the simulation are elliptical galaxies. They look like spheres of stars. These are the most common galaxies found in clusters of galaxies. Elliptical galaxies are different from our galaxy which is a disk galaxy. It looks like a disk of stars.

I am guessing, in reference to your second point, you are referring to the fate of the universe. The simulations here are on a much smaller scale than the universe.

[Emn1ty]

Well... maybe the humans on that star live more millions of years :P. I was just joking though. Besides... the chances of there even being life on one of the planets orbiting one of those stars are even greater than astronomical.

[Emn1ty]

Wouldn't want to be on a planet orbiting one of those stars that get ejected. Bad days follow there. Cool little simulation though, interesting to watch this kind of stuff.

[rberring]

Actually you probably wouldn't notice anything except that there would be fewer stars in your nighttime sky. Also thanks for the comment.

[rberring]

I should probably add that nothing would happen within the lifetime of a human. It would take ~100 million years for the planet to leave a galaxy from which it is ejected. Yes you would be speeding from the galaxy at a speed greater than ~300 kilometers per second (imagine traveling from New York, NY to San Francisco, CA in about 10 seconds), but the distances are so vast that within the lifetime of a human, you would notice no change.

[Jasnator]

is it true the chances of two stars colliding with each other upon galaxy merging is next to 0%?

[rberring]

Yes, that is correct. When two galaxies merge, the chances of a collision occuring between two stars is small (~0%).

[BrunoTheQuestionable]

Thanks for the contribution to the Astronomy group.

Its amazing how significant gravity is at this scale !

[rberring]

Actually the gravitational force dominates on this scale. With the other forces (electromagnetic, weak, and strong) there are negative charges or distance restrictions that prevent them from dominating. With gravity there is no negative mass, and no distance restriction. Thus it dominates on large scales where the other forces dominate on smaller scales.

[BrunoTheQuestionable]

Yes. Even though the electric force is 10^36 times stronger than gravity (for the proton), when the number of particles gets big enough, statistics wins and electromagnetism becomes insignificant compared to gravity. I hadn't appreciated the importance of the statistical factor until now.

[rberring]

Yes. Despite it being the weakest of the four fundemential forces it dominates on the largest of scales. Only in nature will you find something as fascinating and elegant as that.