Why is dark matter distributed in a web rather than falling into clumps ?

[Robin]

Why is dark matter distributed in a web rather than falling into clumps ?
Under the action of gravity normal matter falls into galaxies, stars, planets etc.
How is it that DM doesn't do the same but instead is in web form ?

 

[zhermes]

Why is dark matter distributed in a web rather than falling into clumps ?
Under the action of gravity normal matter falls into galaxies, stars, planets etc.
How is it that DM doesn't do the same but instead is in web form ?

Dark matter does form clumps (e.g. galaxies, galaxy clusters, etc). The exact degree of 'clumpiness' is unknown (especially in regards to streams). It is harder to make dark matter 'clump' compared to baryonic matter, because baryonic matter interacts electromagnetically---thus there exists far more mechanisms to dissipate kinetic energy. Dark matter (to our knowledge) can only redistribute KE via collision-like interactions (e.g. through mergers, tidal/stream interactions, etc).

 

[harcel]

I do not agree that we do not know the clumpiness of dark matter. I think we know this very well from high resolution dark matter simulations.

The cosmic web is a hierarchy of collapsing structures. Overdense regions start to collapse first in one dimension (the direction along which the gradient of the gravitational potential is largest). This forms a 'sheet' which will then contract into one of its two remaining directions to form a filament. The last stage will be the contraction along the filament to form what we call a halo.

In reality, of course, these 'three stages' do not wait for each other to finish, and occur simultaneously. It is just that in some direction this collapse goes faster, due to the stronger gradients in density. This is why we find massive halos at the intersection of filaments of the web.

 

[zhermes]

I do not agree that we do not know the clumpiness of dark matter. I think we know this very well from high resolution dark matter simulations.

A theory---especially one like dark matter---is only as good as its consistency with observations. There are no observations with which to compare the 'clumpiness' of DM. Additionally, even the highest resolution simulations only resolve DM to the order of 10,000 solar masses (e.g. Via Lactea; while some newer simulations have higher mass resolution, their simulation size is too small to examine the effects of tidal and stream interactions and instabilities). It would be absurd to expect high accuracy for arbitrary emergent phenomena (e.g. 'clumpiness') when the simulations are ignorant to the underlying cause (i.e. the specific particle or interactions of DM).

 

[harcel]

The clumpiness of dark matter is not necessarily related to its constituent particles. If you care about those (which is perfectly fine), then observations of the large scale structure of the universe and of galaxy evolution are not going to help you.

Nevertheless, there _are_ observational tests for the clumpiness of DM at the scales you mention (more than roughly 10^4 Msun), from dynamics, dynamical friction, the stirring up of matter due to the movement of clumps of DM.

Clumpiness should be defined at a given density, or all stuff at densities below some value. That is the only way something reasonable will come out. With simulations of given resolution it is straightforward to calculate up to what density you can meaningfully do so. At those scales there are few problems when comparing simulations to observations, but in my humble opinion those comparison are as yet quite basic.

 

[Robin]

The last stage will be the contraction along the filament to form what we call a halo.

The filament was my main area curiosity. Whe you speak of contration alonmg the filament, do you mean the DM in the filament falling to the clumps at each end or in towards the axis of the filament ?
My guess is that as the diameter of the filament contracts then the attraction towards the axis of the filament increases way beyond the attraction towards the clumps at the ends. Meaning little falls towards the clumps at the ends. The filament will simply collapse towards the centre of the axis. Conjures up an image ultimately of a black hole formed along a line rather than a point !
The other problem I have is if dark matter falls into a black hole & DM doesn't undergo friction as no electrostatic interaction occurs, then it could oscillate for ever from one side to the other inside the black hole.

 

[Chronos]

Large scale DM distribution is a consequence of quantum fluctuations in the big bang. DM filaments are merely a magnified view of overdense regions of the BB. Expansion stretched out these regions into the network of filaments now observed.

 

[harcel]

No, that is not how it works. Whether it were quantum fluctuations or whatever else also does not matter. The expansion is not responsible for the formation of filaments, gravity is.

As I explained in more detail above, overdense regions contract, first along one axis, forming sheets, then along the next, forming filaments, followed by the last major axis to form halos. One filament usually ends in very massive halos, but smaller halos also form in the filaments. So it basically looks something like a string (the filament), which is cut in pieces. All these pieces contract to form halos.

 

[Chronos]

Gravity is a secondary effect. DM filaments are magnified views of the original distribution. Overdense regions were held together by gravity.

 

[harcel]

No. Gravity is the primary, and almost only effect. The expansion plays almost no role. The only reason we need it is to make the structures now as big as we see them, they would be a lot smaller, but present, without expansion.

In the early none of the overdense regions (at least almost none) where gravitationally bound. Gravity increased the density contrast, such that the overdense regions could collapse at some point and decouple from the expansion.

When we do simulations of the evolution of the large scale structure, we do not even include expansion explicitly. We use coordinates that grow along with the expansion (so called comoving coordinates), so if the expansion would be different the scales of everything in the universe would be different, but the density contrasts (set purely by gravity) would be identical.