Understanding neutron star mergers
Complex numerical simulation sheds light on an extreme cosmic process
Scientists from the Max Planck Institute for Gravitational Physics in Potsdam and from the Universities of Kyoto and Toho have succeeded for the first time in studying the entire process of two neutron stars orbiting and merging with each other in a long numerical-relativistic simulation. Until now, only simulations describing about 0.1 seconds of the entire process were feasible for such binary systems. The new modeling, which took the Japanese high-performance computer Fugaku 200 days to compute, maps a time span ten times longer and provides new insights into the formation of heavy elements. The study has now appeared in the journal Physical Review Letters.
When in August 2017 gravitational waves from two merging neutron stars were detected, it was a scientific sensation and the beginning of multimessenger astronomy, which combines measurements from gravitational-wave detectors with observations from telescopes that pick up signals in the electromagnetic range. However, it is still not known what exactly happens during and after such a merger. To learn more about neutron stars, which typically have 40% more mass than our Sun at a diameter of about 20 kilometers, highly accurate theoretical computations are needed. The two neutron stars in the simulation published today had 1.2 and 1.5 solar masses, respectively, which is consistent with the parameters of the merger observed in August 2017.
Numerical-relativistic simulation of a binary neutron star merger
Scientists expect to observe coalescing neutron stars also during the recently started fourth observation run of the gravitational-wave detectors. For the interpretation of such signals, reliable theoretical predictions are crucial, which are now available for the first time. "Until now, it has always been necessary to combine different methods to model the complete process of the inspiral, merger and the post-merger phase," explains Kenta Kiuchi, group leader in the Computational Relativistic Astrophysics Department at the Max Planck Institute for Gravitational Physics and first author of the paper. "In addition, previous studies included a lot of assumptions that were not always physically motivated. Our study, on the other hand, is self-consistent and making only a few assumptions. It is the first complete computation of the entire process and provides an accurate picture of the mass ejection during and shortly after the binary neutron star merger."
The formation of heavy elements
The study took 72 million CPU hours on the Japanese high-performance computer Fugaku to simulate one crucial second of the entire process: the last five orbits, the merger itself, and the phase afterwards. "With this long simulation, we have learned a lot about the physics of neutron star coalescences," says Masaru Shibata, director of the Computational Relativistic Astrophysics Department. "It's becoming more and more clear that the elements heavier than iron are in fact synthesized in such extremely energetic processes when matter is ejected from the system during and after the merger."
One second-long numerical-relativistic simulation of a binary neutron star merger
The researchers closely studied the ejection of mass from the system and found that matter is ejected starting about 10 milliseconds after the merger. After 40 milliseconds, this dynamic mass ejection peaks, then flattens out, and about 300 milliseconds after the merger, matter is again ejected – this time from the torus that formed during the merger. While the dynamic mass ejection is due to tidal forces and shock heating during the merger, ejection of matter after the merger results from turbulence in the torus as the scientists have now been able to self-consistently show for the first time.