The recent discovery of GW150914 , the binary black hole merger detected by Advanced LIGO , has the potential to revolutionize observational astrophysics .
But to fully utilize this new window into the universe , we must compare these new observations to detailed models of binary black hole formation throughout cosmic time .
Expanding upon our previous work [ ] , we study merging binary black holes formed in globular clusters using our Monte Carlo approach to stellar dynamics .
We have created a new set of 52 cluster models with different masses , metallicities , and radii to fully characterize the binary black hole merger rate .
These models include all the relevant dynamical processes ( such as two-body relaxation , strong encounters , and three-body binary formation ) and agree well with detailed direct N -body simulations .
In addition , we have enhanced our stellar evolution algorithms with updated metallicity-dependent stellar wind and supernova prescriptions , allowing us to compare our results directly to the most recent population synthesis predictions for merger rates from isolated binary evolution .
We explore the relationship between a cluster ’ s global properties and the population of binary black holes that it produces .
In particular , we derive a numerically calibrated relationship between the merger times of ejected black hole binaries and a cluster ’ s mass and radius .
With our improved treatment of stellar evolution , we find that globular clusters can produce a significant population of massive black hole binaries that merge in the local universe .
We explore the masses and mass ratios of these binaries as a function of redshift , and find a merger rate of \sim 5 Gpc ^ { -3 } yr ^ { -1 } in the local universe , with 80 % of sources having total masses from 32 M _ { \odot } to 64 M _ { \odot } .
Under standard assumptions , approximately 1 out of every 7 binary black hole mergers in the local universe will have originated in a globular cluster , but we also explore the sensitivity of this result to different assumptions for binary stellar evolution .
If black holes were born with significant natal kicks , comparable to those of neutron stars , then the merger rate of binary black holes from globular clusters would be comparable to that from the field , with approximately 1 / 2 of mergers originating in clusters .
Finally we point out that population synthesis results for the field may also be modified by dynamical interactions of binaries taking place in dense star clusters which , unlike globular clusters , dissolved before the present day .