intense and dense afternoon as we get the current state of the art about neutron star natal kicks, theory and observational constraints
need >~ 2.1-2.3 solar mass He cores to get to iron and classic type II SNe
stable mass transfer may lower this limit
electron capture supernovae, in binaries, may limit on core masses right before SNe town to ~ 1.45 solar masses or more
evidence other than presence in globulars for subpopulation of neutron stars having low kick speeds: high mass x-ray binaries with e 30 days
- unlikely to have had strong tidal evolution, so the NS in these systems look like they have low SNe kicks, of maybe
there are other ways to get formation channels with low kicks - Fe core collapse but maybe with rapidly rotating cores, or low mass cores with stripped envelopes, or, what the heck, maybe it is the dratted magnetic field...
use the known NS-NS systems in the Milky Way, for which there is enough information - masses, spin orientation and system velocity
to constrain kicks on the second neutron star formed in these systems.
oh boy, biases when marginalizing 5-D probability distribution functions
there is a reason why these are not often discussed... read the paper!
bottom line, integrating PDFs by eye is not a robust process for typical astronomers...
so PSR0737-3039 may have had second NS through electron capture supernova with associated low velocity natal kick
1534 and 1913 are consistent with high kick Fe core collapse from higher mass core
summarises why NS with natal kicks are retained is a problem, very nicely
get ~ 10% retention fraction with electron capture supernovae, for suitable range of binaries and mass ranges, at low Z, that go to ECS channel
NB: ECS formed NS have lower gravitational mass, so expect 80-90% of NS in globulars to be relatively low mass (