AGN jets are one of the most powerful phenomena in our universe. They seem to significantly influence the growth of galaxies/clusters through AGN feedback and they might be the primary source of extragalactic cosmic rays. Thus it is important to understand how jets accelerate and how their energy balance evolves from launch near the black hole to 5-8 orders of magnitude in distance where these jets interact with the ambient medium. Taking all physical effects including time dependence into account is challenging on the theoretical side. That’s why numerical simulations are used to complement other analytical work. However, since these simulations span at least 5 orders of magnitude in distance they are computationally very challenging.
To address these and other questions I developed a GPU version of the GR-MHD HARM code that is able to evolve a full disk-jet system in high resolution for over 5 orders of magnitude in distance using axisymmetry on a single GPU. After describing some of the computational work I performed, I will discuss these simulation’s results and their physical implications. These full disk-jet simulation results do not seem to fully agree with more idealised simulations, where the jet’s shape is fixed via an external ‘wall’ and thus some instabilities at the jet-wind interface are suppressed. I will eventually finish my talk by giving a glimpse into my future work regarding the causal structure of jets, which according to theoretical work may be very important in explaining standing shock features in AGN jets.