New composite models of partially ionized protoplanetary disks

We study an accretion disk in which three different regions may coexist: MHD turbulent regions, dead zones and gravitationally unstable regions. Although the dead zones are stable, there is some transport due to the Reynolds stress associated with waves emitted from the turbulent layers. We model the transport in each of the different regions by its own $\alpha$ parameter, this being 10 to $10^{3}$ times smaller in dead zones than in active layers. In gravitationally unstable regions, $\alpha$ is determined by the fact that the disk self-adjusts to a state of marginal stability. We construct steady-state models of such disks. We find that for uniform mass flow, the disk has to be more massive, hotter and thicker at the radii where there is a dead zone. In disks in which the dead zone is very massive, gravitational instabilities are present. Whether such models are realistic or not depends on whether hydrodynamical fluctuations driven by the turbulent layers can penetrate all the way inside the dead zone. This may be more easily achieved when the ratio of the mass of the active layer to that of the dead zone is relatively large, which in our models corresponds to $\alpha$ in the dead zone being about 10% of $\alpha$ in the active layers. If the disk is at some stage of its evolution not in steady-state, then the surface density will evolve toward the steady-state solution. However, if $\alpha$ in the dead zone is much smaller than in the active zone, the timescale for the parts of the disk beyond a few AU to reach steady-state may become longer than the disk lifetime. Steady-state disks with dead zones are a more favorable environment for planet formation than standard disks, since the dead zone is typically 10 times more massive than a corresponding turbulent zone at the same location.