Large Scale Simulations of Jets in Dense and Magnetised Environments

We have used the vectorised and parallelised MHD code NIRVANA on the NEC SX-5 in parallel mode to simulate the interaction of jets with a dense environment on a scale of more than 200 jet radii. A maximum performance of 0.75 GFLOP per processor could be reached. One simulation is axisymmetric and purely hydrodynamic, but with a resolution of 20 points per beam-radius (ppb). The bipolar jet is injected in the center of a spherically symmetric King profile, initially underdense to its environment by a factor of 10,000. As expected from our previous work, the jet starts with producing a spherical bubble around it, bounded by the bow shock. The bubble slowly elongates, first with roughly elliptical shape, and then forms narrower extensions in beam direction. The final aspect ratio of the bow shock is 1.8. We have transformed the results on a 3D-rectangular grid and integrated the emission properties to compare the results with observed central cluster radio galaxies. In the particular case of Cygnus A, we come to convincing consistency, morphologically, regarding the size of the influenced region by the jet, size, and cylindrical shape of the radio cocoon, and source age. This strongly supports our earlier hypothesis on the nature of the jet in Cygnus A, and the derived constraints on other jet parameters like a power of 8×10 46 erg/s and an age of 27 Myr. But, the simulation also clearly shows the shortcoming of the model: The jet's beam is very unstable, reaching the tip of the bow shock only very seldom. Also, the contact discontinuity between shocked beam plasma and shocked ambient gas is quite disrupted by the action of the Kelvin-Helmholtz-instability. This is not seen in observations, and necessitates the presence of dynamically important magnetic fields or an at least moderately relativistic flow, or both. The other simulation was designed to explore the impact of a jet on a randomly magnetised environment. A bipolar jet is injected into a King profile on a 3D Cartesian grid, with a resolution of 3 ppb. This simulation was not successful, because the timestep became too low after 0.7 Myr The preliminary results show no increase of the magnetic fields reversal scale yet, which was expected from an unpublished 2D slab jet simulation, but not that early.