High-accuracy drivers to simulate black hole binaries beyond general relativity with the fixing-the-equations approach

Scientific Impact Assessment

Core Contribution. This paper presents a detailed methodological advance for simulating black hole binaries in theories beyond general relativity, specifically shift-symmetric scalar Gauss-Bonnet (sGB) gravity, using the "fixing-the-equations" approach implemented in the `spectre` pseudo-spectral discontinuous Galerkin code. The central novelty is a new family of "comoving driver equations" that exploit the approximate helical symmetry of quasicircular binaries so that the auxiliary variables recover the exact (quasi-)stationary solutions of the fully-coupled theory throughout the inspiral. A second, more subtle contribution is the identification of a genuine pathology: treating tensor auxiliary variables component-wise as scalars (as done in prior literature) induces spurious growth of BH spins over long timescales. The authors trace this to improper frame-transformation of tensor components and fix it via Lie-dragging (the CD2 driver). The result is unusually long (~40 GW cycles) and accurate (phase error ≲O(1) rad) beyond-GR waveforms, incorporating memory and extracted at null infinity with CCE.

Methodological Rigor. The work is methodologically strong. Single-BH results are validated against analytic scalar-charge predictions with clear finite-radius error scaling. Convergence is demonstrated in both numerical resolution (p-refinement) and driver timescale, with careful monitoring of constraint energy and tracking diagnostics. The authors directly compare against the closest prior work (Paper B / Corman et al.) and against alternative driver formulations (advective, wavelike) in appendices, showing quantitative improvements of several orders of magnitude in tracking. They are commendably candid about limitations: the use of GR initial data (not equilibrium sGB data) causes transients; the ambiguity of applying the Christodoulou mass formula in sGB; the empirical lower bound σ≳ℓ² limiting achievable accuracy. Appendices provide detailed first-order system and source-term expressions. The appendix (B) explicitly checking for the secular amplitude growth that plagued the earlier reduction-of-order approach is a strong control.

Potential Impact. The subfield — inspiral-merger-ringdown NR simulations beyond GR — is small but strategically important given next-generation detectors (Einstein Telescope, Cosmic Explorer, LISA) and the drive toward theory-agnostic and theory-specific tests of GR. Prior beyond-GR waveforms were limited to a few orbits with significant secular errors. This work meaningfully raises the bar on length, accuracy, and robustness, and lays groundwork for calibrating the first full beyond-GR EOB waveform model (Julié et al.) that could enable Bayesian model selection between GR and sGB on real events. The comoving/Lie-derivative driver design and the tensor-pathology warning are reusable insights applicable to other theories treatable by fixing-the-equations (EFT extensions, k-essence, vector theories). This is likely to be cited and built upon by the handful of groups doing beyond-GR NR.

Timeliness & Relevance. Highly timely. Multiple groups have recently achieved beyond-GR merger simulations; the explicitly stated "next challenge" is exactly robustness, length, and accuracy — which this paper targets. The identification of a spin-growth artifact that could contaminate phase measurements is directly relevant to ongoing efforts.

Strengths.

Limitations / Weaknesses.

Other observations. The paper functions substantially as a foundational methods/tools contribution supporting a companion Letter. Reproducibility is aided by the open `spectre` code and thorough appendices, though the practical barrier to replication (compute, code familiarity) is very high. The refutation dimension is nontrivial: the authors show a previously-used tensor-driver prescription is inadequate for long evolutions — a useful corrective, framed carefully rather than sensationally. The corroboration of Ripley-Pretorius's negative-area observation and the σ≳ℓ² driver-scale limit provide independent confirmation of prior findings.

Overall, this is high-quality, rigorous, specialist work that advances the state of the art in a narrow but consequential subfield. Its impact will be real but concentrated within the beyond-GR NR and GW-testing communities rather than broadly transformative.