Previous
abstract 
Next
abstract
We investigate the interaction between a stellar wind that has been equatorially directed due to rotation and a pre-existing Keplerian circumstellar disk. Above a threshold rotation rate, the stellar wind forms a shock-compressed zone adjacent to the Keplerian disk. Within this zone there is simultaneous infall (in the inner region) and outflow (in the outer region. We consider two regimes for the interaction between the wind and disk material --- entrainment or complete mixing. Entrainment occurs when the wind flows across the surface of the disk with a turbulent boundary layer. In the inner recirculation zone, the entrained disk material is accreted onto the star, while in the outer zone, it is carried outward with the wind. Using results for mixing layers in stellar outflows, we estimate the length of the entrainment region. When this region is very small, we instead consider the second regime of complete mixing of the wind and disk material. In this case, the addition to the disk of low specific angular momentum wind material results in accretion of the inner region of the disk, clearing a hole. In the outer region, the equatorially concentrated wind then ablates the inner edge of the disk, carrying away the outer disk material. We find that the clearing time is given by the ratio of the disk mass to stellar mass loss rate. For Be stars, this timescale is less than a year, indicating that a Keplerian disk cannot persist around a Be star unless it is constantly replenished. Applying the same mechanism to Herbig Ae/Be stars after the disk accretion has decreased below the stellar mass loss rate, we find a clearing time of $10^6$ years; thus, wind induced accretion may play a role in removing the disks around massive young stars. Finally, it has been suggested that the mid-plane ring around SN1987a may be the result of a hole cleared by photoevaporation of the remnant protostellar disk. Interestingly, the timescale to clear a hole out to the radius of the ring by wind induced accretion/ablation is of order $10^7$ years, so this mechanism may also be a viable candidate for producing the midplane ring.
