Subhaloes in Self-Interacting Galactic Dark Matter Haloes (paper)
Mark Vogelsberger, Jesus Zavala, Abraham Loeb
A series of N-body simulations (so gravity only, no worrying about computationally expensive, or indeed theoretically poorly understood gas and stellar physics) of objects that are of similar total mass to the Milky Way halo. So far so Aquarius (which indeed this paper uses) but the nice take on the problem is that the Dark Matter is assumed to self-interact. There's no theoretical reason why it shouldn't (and indeed they reference a Yukawa-like gauge boson interaction that might have just such a velocity-dependent interaction cross-section) but that's beyond my area of expertise, besides it's not a new idea so feel free to wiki it probably. Instead all we need to know is that this could happen and if so, what are the consequences of Dark Matter that can?
First off I really like the way they implemented this into the simulation, very clever and in principle very simple. Just work out the probability of interaction with the nearest 38 neighbours weighted by both the 3D cubic spline kernel and the interaction cross-section rate (both modelled as a constant, which is ultimately strongly disfavoured and also one which depends on the relative velocities of the particle and its neighbours). Then just add (and subtract) half the relative velocity to the particle (and neighbour) to the centre of mass velocity of the two at a random direction... (uniform elastic scattering basically).
As shown in their work the density profile forms a 'core' of about 1 kpc (3000 lightyears or so across) rather than rising to infinite density. However the profile outside of their region is in agreement with vanilla Cold Dark Matter. The satallites around the 'Milky Way' are in less concentrated in better agreement with the dwarf spheroidals observed around our own galaxy which was a nice result (typically CDM substructure is too dense, i.e. has too high a circular velocity at a given radius...).
The biggest surprise (to me) is that the mass function of the subhaloes doesn't change for the velocity dependent models (only the constant cross-section model, which is already ruled out due to conflicts with shapes of galaxy clusters). I would naively have thought that low density cored structures would have a lower survivability rate than dense infalling nuggest. In any case the 'missing satallite' problem remains where more DM subhaloes are predicted to orbit the galaxy than have been detected as satallites...
So in conclusion, some really nice work, a fun new take (as far as I was aware) on Dark Matter but I'll still go with galaxy formation physics explain the discrepancies (although I still think a luke-warm Dark Matter candidate could also help!).