= Solution
Without feedback, efficiently cooling gas would turn into stars too readily. Young stars return radiation, winds and <supernova> <energy> and <momentum> to their surroundings. This can heat gas, disrupt dense clouds and drive outflows, with the strongest fractional effect in shallow potential wells. The outflow <mass-loading factor> $\eta_w=\dot M_{\rm out}/\dot M_*$ is useful, but its dependence on escape speed differs between energy-limited and momentum-limited models. Some expelled gas later recycles; feedback also transports metals and changes subsequent cooling.
<active-galactic-nucleus feedback> is important in massive galaxies and hot haloes. Radiative or mechanical power from <black-hole accretion> can suppress the supply of cold gas. Collimated jets inflate radio/X-ray cavities, while broader winds can carry significant <momentum> and kinetic <energy>. For a relativistic cavity fluid with adiabatic index $4/3$, its <enthalpy> is approximately $4pV$, so $4pV/t_{\rm age}$ estimates mechanical power. Comparing it with the gas cooling luminosity tests whether that <energy> budget is sufficient, rather than merely showing that a radio source is present.
Deep Chandra observations of Perseus resolved bubbles, a density jump and weak surface-brightness ripples consistent with repeated mechanical disturbance of the intracluster gas. Their interpretation as heating requires a dissipation mechanism and a coupling history, not just a large instantaneous power.<A deep Chandra observation of the Perseus cluster: shocks and ripples by A. C. Fabian and others>[abstract; density front, radio bubbles and the proposed heating/cooling balance]{ref}
Observed outflow geometry, filling factors, cavity ages and cooling corrections create substantial uncertainties. Feedback can also compress gas and locally promote <star formation>. Simulations must track radiative losses and the scale on which <energy> or <momentum> couples: depositing unresolved thermal <energy> is not automatically equivalent to driving a resolved wind. A successful model should explain stellar <masses>, gas fractions, <metallicity> and circumgalactic or intracluster thermodynamics together.
Back to article page
