The speed of the 'bullet' in the merging galaxy cluster 1E0657-56

Deep Chandra exposures of the hot galaxy cluster 10657-56 have revealed a highly dynamical state of the system due to an ongoing merger with a massive subcluster. The system is observed shortly after the first core-passage of the infalling subcluster, which moves approximately in the plane of the sky and is preceded by a prominent bow shock with Mach number M similar to 3. The inferred shock velocity of similar to 4700 km s(-1) has been commonly interpreted as the velocity of the 'bullet' subcluster itself in the rest frame of the parent cluster. This velocity is unexpectedly high in the Lambda cold dark matter (ACDM) cosmology, which may require non-trivial modifications in the dark sector such as additional long-range scalar forces if taken at face value. Here we present explicit hydrodynamical toy models of galaxy cluster mergers which very well reproduce the observed dynamical state of 1E0657-56 and the mass models inferred from gravitational lensing observations. However, despite a shock speed of 4500kms(-1), the subcluster's mass centroid is moving only with similar to 2700kms(-1) in the rest frame of the parent cluster. The difference arises in part due to a gravitationally induced inflow velocity of the gas ahead of the shock towards the bullet, which amounts to similar to 1100 km s(-1) for our assumed 10: 1 mass ratio of the merger. A second effect is that the shock front moves faster than the subcluster itself, enlarging the distance between the subcluster and the bow shock with time. We also discuss the expected location of the lensing mass peak relative to the hydrodynamical features of the flow, and show that their spatial separation depends sensitively on the relative concentrations and gas fractions of the merging clusters, in addition to being highly time dependent. A generic ACDM collision model, where a bullet subcluster with concentration c = 7.2 merges with a parent cluster with concentration c = 3 on a zero-energy orbit, reproduces all the main observational features seen in IE0657-56 with good accuracy, suggesting that I E0657-56 is well in line with expectations from standard cosmological models. In theories with an additional 'fifth' force in the dark sector, the bullet subcluster can be accelerated beyond the velocity reached in ACDM, and the spatial offset between the X-ray peak and the mass centroid of the subcluster can be significantly enlarged. Our results stress the need for explicit hydrodynarnical models for the proper interpretation of actively merging systems such as 1E0657-56.