CERN Accelerating science

 
Colliding clusters and dark matter self-interactions - Kahlhoefer, Felix et al - arXiv:1308.3419CERN-PH-TH-2013-194NSF-KITP-13-140CP3-ORIGINS-2013-029DIAS-2013-29OUTP-13-14P
 
Collision between two DM particles in the rest frame of one of the particles (left) and in the centre-of-mass frame (right).
The expected separation between DM halo and galaxies as a function of the distance between sub-cluster and main cluster for different definitions of the two respective populations. The green dot-dashed line indicates the separation that results when all particles initially bound to the sub-cluster are included in the centroid calculation. For the blue dotted (purple dashed) lines only the inner $80\%$ ($68\%$) of the two distributions are included. For the solid magenta line, we additionally exclude those regions where the surface density of the main cluster dominates. Note the important differences between the case of frequent self-interactions (top panel) and rare self-interactions (bottom panel). System A is representative of Abell 520.
One-dimensional distribution of DM particles (blue) and galaxies (orange) along the line of motion of the DM sub-cluster for System A (representative of Abell 520) and System B (representative of the Bullet Cluster) for frequent self-interactions. The relative normalisation of the two distributions has been chosen in such a way as to facilitate comparison. One can clearly distinguish the tail of galaxies which are moving ahead of the DM halo.
noimgSummary of the parameters used in the numerical simulations. All clusters are modelled using Hernquist profiles. See Appendix~\ref{ap:numerical} for details. System A is representative of Abell 520, System B of the Bullet Cluster and System C of the Musket Ball Cluster.
One-dimensional distribution of DM particles (blue) and galaxies (orange) along the line of motion of the DM sub-cluster for System A (representative of Abell 520) and System B (representative of the Bullet Cluster) for frequent self-interactions. The relative normalisation of the two distributions has been chosen in such a way as to facilitate comparison. One can clearly distinguish the tail of galaxies which are moving ahead of the DM halo.
Observed separation resulting from frequent DM self-interactions for various DM self-interaction cross-sections as a function of the cluster distance for System A (representative of Abell 520), System B (representative of the Bullet Cluster) and System C (representative of the Musket Ball Cluster). The bottom-right panel also shows for System C the separation as a function of the cross-section for various distances. Note that a self-interaction cross-section of $\tilde{\sigma} / m_\text{DM} = 1.6\,\text{cm}^{2}\,\text{g}^{-1}$ implies an evaporation of up to $40\%$ of the total sub-cluster mass in the Bullet Cluster, in slight tension with observations.
Two-dimensional distribution (i.e.~surface density) of DM after the sub-cluster has passed through the main cluster for System A (representative of Abell 520) and System B (representative of the Bullet Cluster) for the case of contact interactions. The black solid contours indicate lines of constant surface density, starting at $10^8\,\mathrm{M}_{\sun}\,\text{kpc}^{-2}$ at the outermost contour and increasing by a factor of 2 with each contour towards the centre. For example, the dark purple region (outermost for both systems) has a surface density of $\Sigma = (\text{1--2}) \times 10^8\,\mathrm{M}_{\sun}\,\text{kpc}^{-2}$, the light yellow region (innermost for System B) has a surface density of $\Sigma > 6.4 \times 10^9\,\mathrm{M}_{\sun}\,\text{kpc}^{-2}$. The green dashed contour indicates the iso-density contour containing $68\%$ of the total halo mass, which is used for the calculation of the centroid (see \S~\ref{sec:separation}).
Two-dimensional distribution (i.e.~surface density) of DM after the sub-cluster has passed through the main cluster for System A (representative of Abell 520) and System B (representative of the Bullet Cluster) for the case of contact interactions. The black solid contours indicate lines of constant surface density, starting at $10^8\,\mathrm{M}_{\sun}\,\text{kpc}^{-2}$ at the outermost contour and increasing by a factor of 2 with each contour towards the centre. For example, the dark purple region (outermost for both systems) has a surface density of $\Sigma = (\text{1--2}) \times 10^8\,\mathrm{M}_{\sun}\,\text{kpc}^{-2}$, the light yellow region (innermost for System B) has a surface density of $\Sigma > 6.4 \times 10^9\,\mathrm{M}_{\sun}\,\text{kpc}^{-2}$. The green dashed contour indicates the iso-density contour containing $68\%$ of the total halo mass, which is used for the calculation of the centroid (see \S~\ref{sec:separation}).
One-dimensional distribution of simulated DM particles and galaxies along the direction of motion of the sub-cluster for the case of contact interactions. The top panel shows three different populations of DM particles: Particles which have scattered over the course of the simulation depending on whether they are still bound (red) or not bound (green) to the sub-cluster and particles which have not scattered during the simulation (blue). The bottom panel shows the sum of these contributions (blue) compared to the distribution of galaxies (orange). The relative normalisation of these two distributions has been chosen in such a way as to facilitate comparison. Note that System A is representative of Abell 520.
One-dimensional distribution of simulated DM particles and galaxies along the direction of motion of the sub-cluster for the case of contact interactions. The top panel shows three different populations of DM particles: Particles which have scattered over the course of the simulation depending on whether they are still bound (red) or not bound (green) to the sub-cluster and particles which have not scattered during the simulation (blue). The bottom panel shows the sum of these contributions (blue) compared to the distribution of galaxies (orange). The relative normalisation of these two distributions has been chosen in such a way as to facilitate comparison. Note that System A is representative of Abell 520.
Observed separation resulting from rare DM self-interactions for various DM self-interaction cross-sections as a function of the cluster distance for System A (representative of Abell 520), System B (representative of the Bullet Cluster) and System C (representative of the Musket Ball Cluster). The bottom-right panel also shows for System C the separation as a function of the cross-section for various distances.
Observed separation in System B as a function of the distance between main cluster and sub-cluster for rare DM self-interactions with $\sigma/m_\text{DM} = 1.2\,\text{cm}^2\,\text{g}^{-1}$ according to our analytical model (solid red line) and our numerical simulation (purple dots). The dashed green (dotted blue) line indicates the contribution to the total separation from particles with $v > v_\text{esc}$ ($v < v_\text{esc}$).