University thesis:
Dissertation, Friedrich-Schiller-Universität Jena, 2023
Footnote:
Tag der Verteidigung: 25.09.2023
Zusammenfassungen in deutscher und englischer Sprache
Description:
The detection and analysis of gravitational waves (GWs) from compact binary systems relies on accurate modeling of the expected signals emitted by such sources. In this thesis we develop computationally efficient yet accurate models for coalescing binary black holes (BBHs) and binary neutron stars (BNSs), relying on the effective-one-body (EOB) framework as implemented in the TEOBResumS family of models. Building on its multipolar aligned-spin avatar, we improve TEOBResumS to include the description of spins precession via an efficient hybrid PN-EOB scheme, thus obtaining a new state-of-the-art inspiral-merger-ringdown model for BBHs and the first multipolar precessing model for coalescing BNSs. We validate our model in terms of NR faithfulness, finding that TEOBResumS agrees to more than 97% with NR results over a considerable portion of the parameter space. Its efficiency is demonstrated by directly employing the model in the parameter estimation (PE) of a handful of events detected by the LVK collaboration (GW150914, GW190412 and GW170817) without the need of surrogates or reduced models. Employing a flavor of TEOBResumS able to model the evolution of systems coalescing along non-circular trajectories, we then study the phenomenology of the GWs that are produced by systems merging along initially unbound orbits. After comparing our waveforms with a set of highly eccentric NR simulations, we analyze GW190521 under the hypothesis that it originated from a dynamical capture of two BHs. Our results suggest that GW190521 may be the the first detected GW signal to correspond to such a system. Finally, we refine the TEOBResumS description of matter effects: after critically assessing the importance of resonant tidal effects for quasi-circular and eccentric BNS mergers, we considerably improve the model performance by including high-order PN information and few NR-informed parameters.