This work in this project is organised through five work packages (WPs). In WPA-1 we develop new numerical techniques that allow for simulations with much higher numbers of stellar particles close to the supermassive black holes. Crucially, we have developed a new version of our KETJU code that allows simulating also supermassive black holes in gas-rich galaxies, which are expected to have strong levels of star formation in their centres. In WPA-2, we develop new models that more accurately describe the accretion of gas onto binary black hole systems, as our runs, unlike previous simulations, are able to dynamically resolve the prolonged binary evolution of supermassive black holes.
In WPs B1-B3, we apply the KETJU code to various astrophysical settings. Specifically, in WPB-1, we have demonstrated that the formation of cores (i.e. regions with low stellar density) in massive galaxies is caused by complex three-body interactions in which the stars are ejected by the binary supermassive black hole. In WPB-2, we have performed for the first time a set of high-resolution cosmological simulations of forming galaxies that include gas physics, star formation, and an accurate description for black hole dynamics. Using these simulations, we have also made predictions for the strength of the gravitational wave signal from merging supermassive black holes in cosmologically forming galaxies. Finally, in WPB-3, we have studied the formation of massive stellar clusters in the mergers of gas-rich dwarf galaxies. Our simulations resolve, for the first time, the internal dynamics of stellar clusters in a global galactic-scale simulation, and we find that stellar clusters can indeed form in such a setting and that they have properties that are in good agreement with the observed population of globular clusters found in the local universe.
The results of this project demonstrate that versatility and applicability of the KETJU code in various simulations, including collisionless (i.e. gas-free) merger simulations and hydrodynamical simulations run in a full cosmological setting. We have demonstrated that the cores of massive early-type galaxies are formed by the scouring effect of interacting supermassive black hole binaries by the ejection of stars. In addition, we have demonstrated the importance of resolving the detailed dynamics of supermassive black holes in cosmological zoom-in simulations, as this will enable the study of triple and multiple systems, not possible in simulations run with traditional softened simulation codes. Finally, we have demonstrated that the central stellar density together with the black hole binary eccentricity are the deciding factors in determining the coalescence time for supermassive black hole binary systems.
The results have been disseminated in a large number of refereed publications. In addition, the results have been presented and discussed in a number of conferences and workshops, with the PI and group members receiving a large number of invitations to present our work. Finally, the research results have been disseminated to the general public through a large number of public talks, media interviews and a few television appearances. In July 2025 we also organised a three-day Workshop in Helsinki on the topics of black hole dynamics, AGN feedback, high-redshift galaxies and star clusters, which lie at the centre of this ERC project.
