A variety of physical heating mechanisms are combined with radiative cooling to explore , via one dimensional hydrodynamic simulations , the expected thermal properties of the intracluster medium ( ICM ) in the context of the cooling flow problem .
Energy injection from type Ia supernovae , thermal conduction , and dynamical friction ( DF ) from orbiting satellite galaxies are considered as potential heating mechanisms of the central regions of the ICM , both separately and in conjunction .
The novel feature of this work is the exploration of a wide range of efficiencies of each heating process .
While DF and conduction can provide a substantial amount of energy , neither mechanism operating alone can produce nor maintain an ICM in thermal balance over cosmological timescales , in stark contrast with observations .
For simulated clusters with initially isothermal temperature profiles , both mechanisms acting in combination result in long-term thermal balance for a range of ICM temperatures and for central electron densities less than n _ { e } \sim 0.02 cm ^ { -3 } ; at greater densities catastrophic cooling invariably occurs .
Furthermore , these heating mechanisms can neither produce nor maintain clusters with a declining temperature profile in the central regions , implying that the observed “ cooling-core ” clusters , which have such declining temperature profiles , can not be maintained with these mechanisms alone .
Supernovae heating also fails to maintain clusters in thermal balance for cosmological timescales since such heating is largely unresponsive to the properties of the ICM .
Thus , while there appears to be an abundant supply of energy capable of heating the ICM in clusters , it is extremely difficult for the energy deposition to occur in such a way that the ICM remains in thermal balance over cosmological time-scales .
For intracluster media that are not in thermal balance , the addition of a small amount of relativistic pressure ( provided by e.g .
cosmic rays ) forestalls neither catastrophic heating nor cooling .
These conclusions are driven largely by the fact that 1 ) DF heating scales approximately as the gas density , while cooling scales as gas density squared , and thus DF heating can not generically balance cooling without fine-tuning ; 2 ) conduction acts to erase temperature gradients , while most observed clusters in fact show strong gradients in the inner regions .
These results strongly suggest that a more dynamic heating process such as feedback from a central black hole is required to generate the properties of observed intracluster media .