Dark Matter Particles Can't Be Collisionless

A NFW halo density profile would be almost universal if dark matter phenomena were attributable to collisionless dark matter particles. But reality doesn't look like what a collisionless dark matter particle model's predictions at all. 

Conventional studies of galaxy clustering within the framework of halo models typically assume that the density profile of all dark matter haloes can be approximated by the Navarro-Frenk-White (NFW) spherically symmetric profile. However, both modern N-body simulations and observational data suggest that most haloes are either oblate or prolate, and almost never spherical.

From a new preprint which proposes a model of dark matter halos that parameterizes this asymmetry.

The observational data, generically, rules out any collisionless dark matter particle explanations for dark matter phenomena if the particles are not very light and wave-like in their behavior. At a minimum, it rules out all collisionless dark matter particle candidates with masses of more than about 10 keV (possibly much lighter, such as fuzzy dark matter models with particles with a mass of 1−2 × 10⁻²³ eV and other ultralight dark matter models).

The asymmetrical dark matter halo preprint doesn't propose a mechanism by which its asymmetry parameter arises; it is a purely phenomenological formula. 

A self-interacting dark matter particle model could help explain this parameter, but still wouldn't explain the tight correlation between the distribution of ordinary matter in a galaxy and the dark matter phenomena in it that is observed, which would require a different kind of fifth force. 

Any theory attempting to describe dark matter phenomena in which a fifth force mediates interactions between dark matter particles and other dark matter particles, and/or between dark matter particles and ordinary matter, of course, has lost ground with Occam's Razor, relative to a theory that only modifies the equations of gravity, as conventionally applied, without resort to any dark matter particles.

On the other hand, Deur's gravity based model (even if it doesn't actually fully replicate weak field general relativity) can naturally address both concerns, without dark matter particles or dark energy. 

In his model, the asymmetrical distribution of ordinary matter is what gives rise to dark matter in the first place, so inferred dark matter distributions should almost never be spherically symmetric, consistent with what we observe. And, in a gravitationally based model, all dark matter phenomena arise from the ordinary matter distribution, so the tight correlation between ordinary matter distributions and the dark matter phenomena in the galaxy is expected rather than a problem.

Deur's model, of course, is not the only one on offer, although it is one of the best that I've seen so far. Some generalizations of MOND that address, among other things, its shortcomings in galaxy clusters, are discussed in another new preprint.