10:00 AM - 10:15 AM
[MGI35-05] Numerical study of the origin of merging binary black holes
Keywords:Black hole, Binary star, Star cluster, Binary population synthesis, Gravitational N-body simulation
Gravitational wave (GW) observatories have discovered many mergers of
black holes (BHs) for these several years. Their origins have been
highly uncertain, and many formation scenarios have been suggested:
isolated binary evolution from population I, II and III stars,
dynamical capture in open clusters, globular clusters, and nuclear
star clusters, and so on. In order to elucidate their origins, we need
to compare properties of observed binary BHs with those of
theoretically predicted binary BHs. Such properties can be merger
rate, primary mass distribution, mass ratio distribution, spin
magnitude distribution, spin-orbit misalignment distribution,
eccentricity distribution, redshift distribution, and so on. Numerical
simulations are powerful tools to investigate properties of binary BHs
theoretically. We have investigated binary BHs formed through isolated
binary evolution and dynamical capture.
Binary population synthesis (BPS) calculations are commonly used to
predict properties of binary BHs. In BPS calculations, we prepare
initial conditions of binary stars: binary component masses,
semi-major axes, and eccentricities. Each star evolves from zero-age
main-sequence stars to stellar remnants like BHs. When either of stars
fill their Roche lobes, mass transfer or stellar merger set in. We
have made stellar evolution tracks of population II and III stars for
the BPS calculations (Tanikawa et al. 2020, MNRAS, 495, 4170). Using
the stellar evolution tracks, we have revealed properties of binary
BHs formed from population III stars. We have found that their merger
rate is 100 times smaller than the observed rate (Tanikawa et
al. 2020, arXiv:2008.01890). However, we have made clear that GW190521
can be formed from population III stars (Tanikawa et al. 2020,
arXiv:2010.0716), where GW190521 is thought to be difficult to be
formed in isolated binaries.
Binary BHs in star clusters are usually studied with gravitational
N-body simulations coupled with BPS calculations. Here, N-body
simulations solve stellar dynamics, and BPS calculations follow single
and binary star evolutions. Using such simulations and detail
three-body simulations, we have predicted merger rate, mass
distribution, mass ratio distribution, spin magnitude distribution,
and spin-orbit misalignment distribution of binary BHs formed in open
clusters. Their merger rate indicates that binary BHs formed in open
clusters can be one of major components of observed binary BHs
(Kumamoto et al. 2019, MNRAS, 486, 3942). Their mass and spin
distributions are also consistent with observed ones' (Kumamoto et
al. 2020, MNRAS, 495, 4268; Trani et al. 2021, submitted).
We are now studying binary BHs formed in globular clusters, massive
star clusters. Gravitational N-body simulations of globular clusters
were impossible, since their calculation costs are extremely high.
Recently, a N-body simulation code ``PeTar'' have been developed. The
PeTar code calculates gravitational forces between all pairs of stars,
reducing their calculation costs with the Barnes-Hut tree
algorithm. Moreover, the PeTar code works well on massively parallel
computing environment. The PeTar code enables us to perform
gravitational N-body simulations of realistically massive and dense
globular clusters. Using the PeTar code, we have shown that a
top-heavy initial mass function does not always increase a merger rate
of binary BHs formed in globular clusters (Wang et al. 2021,
arXiv:2101.09283).
black holes (BHs) for these several years. Their origins have been
highly uncertain, and many formation scenarios have been suggested:
isolated binary evolution from population I, II and III stars,
dynamical capture in open clusters, globular clusters, and nuclear
star clusters, and so on. In order to elucidate their origins, we need
to compare properties of observed binary BHs with those of
theoretically predicted binary BHs. Such properties can be merger
rate, primary mass distribution, mass ratio distribution, spin
magnitude distribution, spin-orbit misalignment distribution,
eccentricity distribution, redshift distribution, and so on. Numerical
simulations are powerful tools to investigate properties of binary BHs
theoretically. We have investigated binary BHs formed through isolated
binary evolution and dynamical capture.
Binary population synthesis (BPS) calculations are commonly used to
predict properties of binary BHs. In BPS calculations, we prepare
initial conditions of binary stars: binary component masses,
semi-major axes, and eccentricities. Each star evolves from zero-age
main-sequence stars to stellar remnants like BHs. When either of stars
fill their Roche lobes, mass transfer or stellar merger set in. We
have made stellar evolution tracks of population II and III stars for
the BPS calculations (Tanikawa et al. 2020, MNRAS, 495, 4170). Using
the stellar evolution tracks, we have revealed properties of binary
BHs formed from population III stars. We have found that their merger
rate is 100 times smaller than the observed rate (Tanikawa et
al. 2020, arXiv:2008.01890). However, we have made clear that GW190521
can be formed from population III stars (Tanikawa et al. 2020,
arXiv:2010.0716), where GW190521 is thought to be difficult to be
formed in isolated binaries.
Binary BHs in star clusters are usually studied with gravitational
N-body simulations coupled with BPS calculations. Here, N-body
simulations solve stellar dynamics, and BPS calculations follow single
and binary star evolutions. Using such simulations and detail
three-body simulations, we have predicted merger rate, mass
distribution, mass ratio distribution, spin magnitude distribution,
and spin-orbit misalignment distribution of binary BHs formed in open
clusters. Their merger rate indicates that binary BHs formed in open
clusters can be one of major components of observed binary BHs
(Kumamoto et al. 2019, MNRAS, 486, 3942). Their mass and spin
distributions are also consistent with observed ones' (Kumamoto et
al. 2020, MNRAS, 495, 4268; Trani et al. 2021, submitted).
We are now studying binary BHs formed in globular clusters, massive
star clusters. Gravitational N-body simulations of globular clusters
were impossible, since their calculation costs are extremely high.
Recently, a N-body simulation code ``PeTar'' have been developed. The
PeTar code calculates gravitational forces between all pairs of stars,
reducing their calculation costs with the Barnes-Hut tree
algorithm. Moreover, the PeTar code works well on massively parallel
computing environment. The PeTar code enables us to perform
gravitational N-body simulations of realistically massive and dense
globular clusters. Using the PeTar code, we have shown that a
top-heavy initial mass function does not always increase a merger rate
of binary BHs formed in globular clusters (Wang et al. 2021,
arXiv:2101.09283).