We have tentative evidence of massive stars that disappear without a bright transient. It is commonly argued that this massive stars have low angular momentum and can collapse into a black hole without significant feedback. In this talk I will make use of general-relativistic hydrodynamical simulations to understand the flow around a newly-formed black hole. I will discuss the angular momentum needed in order for the infalling material to be accreted into the black hole without forming a centrifugally supported structure, thus generating no effective feedback. If the feedback from the black hole is significant, the collapse can be halted and, as a result, it is likely followed by a bright transient. With the results from the simulation, I will constrain the maximum rotation rate for the disappearing massive progenitors know, and set a limit on the rate of expected disappearing stars.