The true role of kinematic viscosity in ICM/radio-galaxy interactions
remains unclear, primarily due to the great uncertainty associated
with magnetic suppression of Spitzer-type transport processes in the
plasma. However, in stark contrast to the inviscid models, our models
of bubble evolution in modestly viscous ICM can reproduce both the
morphology and inferred flow patterns seen around the best studied of
the ghost ICM cavities, the NW cavity in Perseus-A. The principal
effect of viscosity is to stabilize the bubble against both RT and KH
instabilities, thereby allowing it to remain intact as it floats
upwards in the ICM atmosphere. The flattening of such bubbles as they
rise is a natural explanation for the elongated morphology of the NW
cavity in Perseus-A. The ``smoke-ring'' like flow pattern below the
NW cavity inferred from the morphology of the H
filaments can
then arise due to minor fragmentation of the rising bubble (leading to
a trailing torus) or genuine vortex shedding.
If the ICM is indeed viscous at the level treated by our models, the subsequent slowing of the evolution timescales for ICM/radio-galaxy interactions may be the solution to the ``shock problem'', i.e. the lack of strong shocks bounding the ICM cavities that are associated with active radio-lobes. In this scenario, radio galaxies such as Per-A would be rather less powerful than previously thought (Heinz, Reynolds & Begelman 1998), inflating their associated ICM bubbles subsonically. If the evolution is sufficiently slow, radiative cooling might allow the ICM shell surrounding the cavity to cool, again in agreement with Chandra observations of Perseus-A.
Ultimately, disentangling the effects of transport processes (viscosity and thermal conduction) and magnetic fields will require further simulation work coupled with detailed observations. On the theoretical side, it is important to initiate a new generation of simulations which incorporate full MHD as well as transport processes. Qualitatively new and rich dynamics becomes possible upon the inclusion of transport processes within MHD. For example, Balbus (2000) discovered the ``magnetothermal instability'' which occurs in any MHD atmosphere with a temperature profile that decreases with height provided there is thermal conduction along the magnetic field lines. In addition, the viscous transport coefficients will be anisotropic in the presence of a magnetic field, and this may lead to yet further complexity in the dynamics. Such instabilities could well be crucial for understanding the dynamics of radio-galaxy/ICM interactions. Observationally, deep (megasecond) Chandra observations of the nearest radio-galaxy/cluster interactions might allow detailed imaging spectroscopy of fluid instabilities associated with the radio-galaxy/ICM interaction. Furthermore, upon the launch of Astro-E2 in 2005, we will be able to directly probe turbulent velocities within the ICM of nearby (bright) clusters through broadening of their X-ray emission lines. With the caveat that full MHD viscous simulations have yet to be done, the presence or absence of turbulence in the ICM, especially in regions well separated from potential driving sources (such as AGN or merging subclusters) could be an important indicator of the presence of viscosity.