Solving the cooling flow problem through mechanical AGN feedback

Unopposed radiative cooling of plasma would lead to the cooling catastro-phe, a massive inflow of condensing gas, manifest in the core of galaxies, groups and clusters. The last generation X-ray telescopes, Chandra and XMM, have radically changed our view on baryons, indicating AGN heating as the balancing counterpart of cooling. This work reviews our extensive investigation on self-regulated heating. We argue that the mechanical feedback, based on massive subrelativistic outflows, is the key to solving the cooling flow problem, i.e. dramatically quenching the cooling rates for several Gyr without destroying the cool-core structure. Using a modified version of the 3D hydrocode FLASH, we show that bipolar AGN outflows can further reproduce fundamental observed features, such as buoyant bubbles, weak shocks, metals dredgeup, and turbulence. The latter is an essential ingredient to drive nonlinear thermal instabilities, which cause the formation of extended cold gas, a residual of the quenched cooling flow and, later, fuel for the feedback engine. Compared to clusters, groups and galaxies require a gentler mechanical feedback, in order to avoid catastro-phic overheating. We highlight the essential characteristics for a realistic AGN feedback, with emphasis on observational consistency.