1998ApJ...506..780B

Beginning from massive binaries in the Galaxy, we evolve black hole-neutron star (BH-NS) binaries and binary neutron stars, such as the Hulse-Taylor system PSR 1913+16. The new point in our evolution is a quantitative calculation of the accretion of matter by a neutron star in a common-envelope evolution that sends it into a black hole. We calculate the mass of the latter to be ∼2.4 M_☉. The black hole fate of the first neutron star can only be avoided if the neutron star does not go through common-envelope evolution. This can be realized if the two massive binaries are sufficiently close in mass, not more than ∼5% apart, so that they burn helium at the same time. Then their common hydrogen envelope is expelled by the tightening in the double He star system, with attendant hydrodynamical coupling to the envelope. In this way, we obtain a rate of ≲10^–5 per yr per galaxy for production of the narrow neutron star binaries, such as Hulse-Taylor 1913+16 or Wolszczan 1934+12. This is in agreement with estimates based on the observed number of such systems extrapolated to the entire Galaxy, with beaming factors and corrections for the ∼90% of binary pulsars estimated to be unobservable. Our chief conclusion is that the production rate for BH-NS binaries (in which the neutron star is unrecycled) is ∼10^–4 per yr per galaxy, an order of magnitude greater than that of neutron star binaries. Not only should this result in a factor of ∼10 more mergings for gravitational wave detectors such as LIGO, but the signal should also be larger. We include some discussion of why BH-NS binaries have not been observed, but conclude that they should be actively searched for.