We present a simple physical model of the central source emission in the M87 galaxy. It is well known that the observed X-ray luminosity from this galactic nucleus is much lower than the predicted one, if a standard radiative efficiency is assumed. Up to now the main model invoked to explain such a luminosity is the ADAF (Advection-Dominated-Accretion-Flow) model. Our approach supposes only a simple axis-symmetric adiabatic accretion with a low angular momentum together with the bremsstrahlung emission process in the accreting gas. With no other special hypothesis on the dynamics of the system, this model agrees well enough with the luminosity value measured by Chandra.

We discuss a special case of formation of axisymmetric shocks in the accretion flow of ideal gas onto a Schwarzschild black hole: when the total energy of the flow is negative. The result of our analysis enlarges the parameter space for which these steady shocks are exhibited in the accretion of gas rotating around relativistic stellar objects. Since keplerian disks have negative total energy, we guess that, in this energy range, the production of the shock phenomenon might be easier than in the case of positive energy. So our outcome reinforces the view that sub-keplerian flows of matter may significantly affect the physics of the high energy radiation emission from black hole candidates. We give a simple procedure to obtain analytically the position of the shocks. The comparison of the analytical results with the data of 1D and 2D axisymmetric numerical simulations confirms that the shocks form and are stable.

We present the simulation of 3D time dependent flow of rotating ideal gas falling into a Schwarzschild black hole. It is shown that also in the 3D case steady shocks are formed in a wide range of parameters (initial angular momentum and thermal energy). We therefore highlight the stability of the phenomenon of shock formation in sub keplerian flows onto black holes, and reenforce the role of the shocks in the high luminosity emission from black hole candidates. The simulations have been performed using a parallelized code based on the Smoothed Particles Hydrodynamics method (SPH). We also discuss some properties of the shock problem that allow its use as a quantitative test of the accuracy of the used numerical method. This shows that the accuracy of SPH is acceptable although not excellent.

Using simple arguments we show that the canonical thin keplerian accretion disks cannot smoothly match any plain advection dominated flow (ADAF) model. By 'plain' ADAF model we mean the ones with zero cooling. The existence of sonic points in exact solutions is critical and imposes constraints that cannot be surpassed adopting 'reasonable' physical conditions at the hypothetical match point. Only the occurrence of new critical physical phenomena may produce a transition. We propose that exact advection models are a class of solutions which don't necessarily involve the standard thin cool disks and suggest a different scenario in which good ADAF solutions could eventually occur.