Abstract
In this paper we develop a waveform model that accurately reproduces the dynamical evolution of intermediate-mass-ratio inspirals, as predicted by the effective-one-body (EOB) model, and which enables us to shed some light on the form of the self-force for events with mass-ratio $1:6$, $1:10$ and $1:100$. To complement this study, we make use of self-force results in the extreme-mass-ratio regime, and of predictions of the EOB model, to derive a prescription for the shift of the orbital frequency at the innermost stable circular orbit which consistently captures predictions from the extreme, intermediate and comparable-mass-ratio regimes.
Introduction
In the era of advanced gravitational wave detectors like LIGO, researchers need accurate "templates"—mathematical models of waveforms—to identify signals from merging black holes. We are currently good at modeling two extremes: comparable-mass binaries (e.g., 1:1 to 1:6) using computationally expensive Numerical Relativity (NR), and extreme-mass-ratio inspirals (EMRIs, e.g., 1:10⁶) using "self-force" theory, which treats the smaller object as a perturbation of the larger one.
However, there is a "no-man's land" in the middle: Intermediate-mass-ratio inspirals (IMRIs), with mass ratios between 1:10 and 1:1000. NR is too slow for these, and EMRI approximations might be too simple. This paper investigates how to accurately model this regime and whether existing self-force data can be "stretched" to fit IMRIs.
Methodology and Model Calibration
The authors develop a flexible IMRI waveform model and use the EOBNRv2 (Effective-One-Body Numerical Relativity version 2) model as their primary benchmark. Since EOBNRv2 was calibrated against NR for mass ratios up to 1:6 and designed to behave correctly in the test-mass limit, it serves as a reliable proxy for "true" physics.
Key methodological steps included:
- Ansatz Construction: Creating a new mathematical framework for orbital frequency and angular momentum flux.
- Calibration: Tuning the parameters of this IMRI model to reproduce EOBNRv2 dynamics with high precision (better than 1 part in 1000). The specific calibration coefficients for different mass ratios are detailed in the table below.
- Self-Force Comparison: Comparing the calibrated model against existing first-order conservative self-force corrections to identify where EMRI-based approximations fail.
- ISCO Derivation: Deriving a new analytic fit for the frequency shift at the innermost stable circular orbit (ISCO) that remains accurate across all mass regimes.
Key Insights: The Self-Force Gap
The most significant finding is that first-order self-force corrections (linear in the mass ratio $\eta$) are insufficient for IMRIs. While they work reasonably well for a 1:100 ratio, they deviate significantly from the expected physics as the mass ratio approaches 1:10, especially near the ISCO.
Furthermore, the EOBNRv2 model, despite being calibrated at low mass ratios, is remarkably consistent with perturbation theory even at a 1:100 ratio. This suggests the EOB framework is inherently better at capturing the transition between regimes than pure perturbation theory.
Results: A Unified ISCO Prescription
The proposed IMRI model reproduces the radial and azimuthal evolution of EOBNRv2 to within 0.1% accuracy across the studied mass ratios (1:6, 1:10, and 1:100). To capture the ISCO frequency shift across all mass ratios, the authors found that a simple linear or quadratic fit in $\eta$ is insufficient; a cubic polynomial is required to bridge the gap between comparable and extreme mass ratios accurately.
The paper provides a new global fit for the ISCO frequency: $$M\Omega_{ISCO}^{fit} = \frac{1}{6\sqrt{6}}(1 + 1.2512\eta - 0.0553751\eta^2 + 5.78557\eta^3)$$ This formula matches self-force predictions for small $\eta$ and EOB/NR predictions for large $\eta$. Using only available linear self-force data leads to a noticeable "dephasing" (timing error) in the waveform near the end of the inspiral for IMRIs.
Conclusion and Future Directions
This study focuses exclusively on non-spinning black holes. Real-world black holes likely spin, which significantly complicates the self-force and the resulting waveforms. While the model matches EOBNRv2, direct comparison with high-resolution NR simulations for $q=1:100$ will be necessary once they become computationally feasible. Future work is also needed to integrate the merger and ringdown phases into a single, physically consistent IMRI model.
Summary updated