# Why does dark matter form walls and filaments

But I want to come at this from a different angle. Like the user asking that other question, I was a bit surprised by the walls, filaments and nodes of the large-scale structure of dark matter:

Intuitively, I might have expected more spherical shapes. I got a bit closer to a satisfactory explanation when I heard about an equivalent way of thinking about the situation. Instead of thinking of overdense regions collapsing (e.g. the Zel'dovich pancakes mentioned in the answer to the question linked above), one can think about underdense regions expanding to form voids/supervoids/etc. The voids are (roughly) spherical, and as they push out and collide, they compress the dark matter into walls and filaments. I'm picturing something like blowing bubbles in soapy water, giving a nice intuitive picture.

Now where I get hung up is that there is an obvious symmetry between thinking about overdensities collapsing and underdensities expanding, but there is an obvious lack of symmetry between the structures resulting from collapse and expansion, for instance why wouldn't we get the opposite case, where the voids are filament shaped and the dark matter forms roughly spherical blobs? I have a feeling that the key is in the strictly attractive nature of gravity, but can't really put my thoughts together coherently. Would be interested to hear if anyone can elaborate a bit on this.

Feel free to get technical in the answers, I have a solid math & physics background to help me interpret. Bonus points if you can draw a nice intuitive picture, though :)

• I thought the filaments are formed because the cold, non-interacting dark matter gets dragged and pulled along by moving massive bodies like galaxies. Jan 16, 2014 at 0:38
• have a look at some cosmic string modeling for dark matter arxiv.org/abs/0805.1060 Jan 16, 2014 at 14:56

I am not super well into the math of DM large scale structure, but I have a little intuitive chirp:

There is nothing in the voids pressing the DM together, it is rather the opposite: the DM is collapsing under its own gravity, which makes the DM stick together wile the rest of the Universe is expanding around it. The bubble picture is deceiving, I usually use this image instead:

It is a screenshot of a YouTube-video of a golf ball smashing a jar of Mayo, taken with a high speed camera. Here, I think the physics is more similar - the mayo is also suddenly "trying" to fill out a much larger space, while a competing force - in this case the stickiness rather than gravity - is holding it together. The result is similar, though...

• Hmmm, now I'm trying to remember if the filament structure requires the expansion of space to form, or whether the same thing occurs idependent of whethr space is expanding. Obviously the physical motivation for the ICs is tied to expansion, but ignoring that... does the same structure form? Jan 16, 2014 at 16:37
• Honestly, I don't know.Obviously, the analogy above is very loose, and I merely posted it because I think it is better than the soap bubble one / which does not necessarily imply that it is good ;-) Jan 16, 2014 at 17:49
• At closer thought; no, of course the filamentary structure of Dark Matter doesn't require an expanding Universe - gravity is gravity, after all. But common for DM and mayo-after-golfball-impact is concentration of matter on a decreasing percentage of the available space... Jan 16, 2014 at 21:20

I came across a very simple and elegant explanation yesterday. It combines simple conservation of energy:

$$E=W+T$$

and the virial theorem (which of course applies to the present scenario):

$$2T+W=0$$

giving the seemingly trivial result:

$$E=-T$$

So as the system moves toward a lower energy state, its kinetic energy necessarily increases, and it gets hot (so for dark matter, gains a higher velocity dispersion). It must collapse.

Gravity is not attractive emanating from objects. It is repulsive emanating from voids, and objects result from congealed light pressed together in black hole whirlpools.

• This is complete nonsense and you can't support any of it with experiment. Jul 22, 2017 at 5:39