Universe - bubbles of Dark Matter?

mikewday

I have seen a few other similar posts but wanted to add my thoughts. Is gravity the space between bubbles of dark matter. Matter that we know and dark matter are like oil and water? Gravity is the force created in the structures between the bubbles. Is that why we see the current structure of the Universe? This thought came to mind after watching a show on the Science channel and then playing in dish washing bubbles afterward with my daughter. The structure between the dish soap bubbles looks a lot the known Universe. Patterns in the natural word often get repeated at larger and smaller levels. Is the Universe just a big bubble foam and gravity/matter just the stuff inbetween? The dark matter is "air" inside the bubbles in the dish soap. Has anyone ever done a computer model to see if this would hold true?

eachus

You are right to notice the similarity, and there may be lots of dark matter in the bubbles in the foam. However, the explanation of the structure doesn't require dark matter.

Everyone, well everyone who reads PhysOrg, has heard of the cosmic microwave background radiation (CMBR). This radiation was originally much higher in frequency and since the universe was much smaller when it was created, it was much more intense at that time.

At that time, about 400 million years after the big bang, matter was getting cool enough that the plasma recombined into atoms (or molecules) which were effectively transparent to the radiation. But it didn't happen everywhere at once. If you have a small bubble of normal matter that doesn't absorb (and reradiate) light, light entering the bubble goes right across and pushes on the other side. That push involves absorption and reradiation. If the reradiation is into the bubble--and eventually it will be, repeat the process.

So any cold bubble in the plasma acts as if it is a radiation source. Note that the remaining plasma is compressed by this radiation, so areas near a bubble are unlikely to form another bubble. When two bubbles collide/merge the process continues, until space is filled mostly with bubbles and most of the matter and radiation is in the sponge part of the foam. As the universe continues to expand, more and more of the radiation is crossing empty bubbles, so the matter rich foam gets to cool down too. But now the matter is much denser, and will act like the plasma absorbing and reradiating the CMBR. This pressure eventually tails off in significance, but in the meantime it has created the large scale structure of the universe.

Where is dark matter in all this? Good question. If dark matter--whatever it is--doesn't absorb (and reradiate) light, most of the dark matter will end up in the voids. If it is subject to gravity, it will get pulled toward the knots in the foam. Note the ifs there--until we know what dark matter is, this is speculation.

mikewday

But would CMBR cause such large structures? Or, is it gravity that takes over and makes the Universe the way we see it today? That was my understanding. I have no background to validate or test any of this but, I think it is interesting to think of gravity as something that is not a force in itself but caused by forces around it such as matter being pushing together by expanding or changing dark matter. I recently saw something that stated that gravity on it's own would not hold a Universe together, it is only with the addition of dark matter into the equation that the Universe stays together. What if gravity is just a side effect of dark matter "bubbles"?

eachus

Very good question. They weren't that big at the time, but what was "that time"?

The problem with simulations is that unless you are very careful, you end up proving what you expected to prove. It is easy to demonstrate that even if the CMBR were perfectly smooth, a foam structure emerges as the universe cools. Much harder is to decide when this process became significant. I think that density variations may have started separating the radiation from the matter at about one million years after the big bang. But I am smart enough to know that even a fairly decent simulation doesn't tell me much, since I get to choose the initial parameters. :-(

Chalnoth

Basically, the way it works is that the dark matter is most of the mass, so when the dark matter clumps, the normal matter tends to fall into the gravitational wells produced by those clumps (forming galaxies, galaxy clusters, etc.). Every time you see a galaxy, imagine it sitting near the center of a much larger clump of dark matter.

Chalnoth

Well, our understanding of the formation of structure in the universe is superb at large scales. The predictions quite precisely match experiment at scales above a few Gpc or so.

It does, however, start to break down when you get to smaller scales. Most especially, CDM models predict many, many more low-mass galaxies than we observe.

The difficulty, however, is that low-mass galaxies, when they first form, tend to lose most of their normal matter because the mass is low enough that the first stars just blow most of the matter out of the galaxy (such low-mass galaxies sometimes have ratios of dark matter to normal matter of a hundred to one or more!). So it may be the case that there are lots of low-mass clumps of dark matter, but they just don't have enough normal matter in them for us to see them.

eachus

A much bigger problem is the shape of the gravity wells created by dark matter. On the scale of kps (kiloparsecs) they have no deep center. Galaxies containing dark matter do have deep gravity wells, but the deep part can be accounted for by the normal matter and supermassive black holes.

Do the black holes contain dark matter, or for that matter, dark energy? Beats me. But Ockham's Razor cautions us, at least, not to assume dark matter flows into black holes. No radiation, so no way to get rid of angular momentum in an accretion disk.

Can the shallowness of dark matter gravity wells be explained the same way? Sounds good, much better than assuming dark matter is in a different brane. But you still need a way to start the dark matter spinning. So we are back to multiple regions where radiation pressure sends (normal) matter in all directions. But if dark matter doesn't interact with electromagnetic radiation, what then? Does the normal matter drag dark matter around via gravity?

Hmm. New theory as of two minutes ago. Start by throwing out neutrino oscillation and replace it with a sea of neutrinos of all flavors, and momentum transfers in collisions between them. (In other words the electron antineutrino and tau antineutrio have a much higher collision cross section than any flavor neutrino with normal matter.) So what we see as neutrino oscillation is really neutrino billiards.

Now we can have neutrinos with mass (well we could have that already). But if the symmetry breaking that gives us a normal matter universe (as opposed to one with equal amounts of matter and antimatter) is also repeated with the higher level particles, then we have a sea of leftover tau and muon anti-neutrinos that can't convert into any lower energy particle, due to conservation of muon and tau lepton numbers. Add a significant mass (order of a dozen MeV/c2) for the tau (anti-)neutrino and something less for the muon (anti-)neutrino. Done.

Chalnoth

No, actually, we can't say that's a problem. I talked with some simulations people about this issue a few weeks back, and it turns out that when you start getting to the center of the gravitational wells, the resolution of the simulation tends to muck things up, and the predictions start to become fuzzy. So we just can't say that our current dark matter models predict a cuspy center.

eachus

Lol! And definitely not disagreeing. You did see my comment about choosing inputs to a simulation? Without real hard numbers about what dark matter is, slight changes in the model inputs result in outputs that support just about any model.

If the simulation says, "No, you are wrong!" great. Falsifying models is the goal. But you learn nothing from simulations, if the best you can say is "I think the initial conditions required to get this result are extreme and unlikely."

The problem is that we are observing the universe from the far end of time. To some extent, you have to tune your model to match the outcomes that we see all around us. But how much is necessary tuning to fit reality, and how much is tuning your model to fit your theory? (For example, do you treat the vacuum energy as constant, or as a function of time? Assuming it varies as one over t cubed seems pretty extreme. Or is it? And do you adjust for inflation, and how and when? Building inflation into your model is just as risky as not doing so.)

I was serious about that neutrino billiards theory. I may need to burn quite a few kilowatt hours to get simulation results, but the model should result in testable predictions. Replacing neutrino oscillations with a sea of all three (anti-) neutrinos at low energies is testable, and even though the model allows for lots of tunable model inputs, assuming that the branching ratios for all three families of particles prior to symmetry breaking are identical should make Ockham happy.

Chalnoth

That's not true, though. As I said earlier, the simulations are only unreliable on small scales. On large scales, the results are highly robust, and match with observation very well.

Also false, because our models of the universe also set the initial conditions.

The main problem with simulations is resolution. While in principle we know most of the physics that goes into the simulations, it's generally not feasible to simulate things all the way down to the level of stars, which is actually necessary to get the right answer from first principles (supernova explosions have large, far-reaching effects on galactic evolution, for instance). So it's not so much a problem of initial conditions as it is a problem that there are a number of complex phenomena going on which are, in principle, known, but we can't actually simulate them. What we need to do, then, is formulate some fudge factors that sort of sweep the underlying complexities under the rug, and then estimate those fudge factors either by observation or with small-scale simulations.

So really, the problem with simulations is not an initial conditions problem. It's a computational problem. It's a problem of finding the right approximations to the physics we know so that the answer is still reliable.

But, at the very least, the large-scale results are independent of these small-scale complexities, and so there is no problem at large scales.