Tidal Disruption and Breakdown of the Worldline Expansion

Tidal disruption is not merely a large correction to a point-particle orbit. It is the loss of a controlled hierarchy: the body’s deformation becomes nonlinear, a surface can shed mass, shocks and streams form, and a single compact worldline may no longer represent the matter. Scale comparisons can warn that this transition is approaching, explicit modes and nonlinear response operators can extend the description, and relativistic hydrodynamics can follow the material through disruption. None of these tools supplies a universal, sharply invariant disruption radius by itself.

Helpful background. The Einstein Equation supplies relativistic gravity; Weak Metric Perturbations and Gauge Freedom supplies the limits of perturbative deformations; Presymplectic Potentials and Currents supplies energy transfer; Relativistic Matter and Compact Objects supplies stellar equilibrium and equations of state; Relativistic Two-Body Dynamics supplies scale-separated worldlines; and Black-Hole Perturbation Theory supplies capture scales around black holes.

Scale hierarchy at the onset of disruption

For a body of mass $m$ and radius $R$, define the electric-curvature scale and its spatial variation scale in the body’s rest frame by

$$\mathcal L_{\mathcal E} =\left(\mathcal E_{ab}\mathcal E^{ab}\right)^{-1/4}, \qquad \mathcal L_{\nabla\mathcal E} \sim\frac{|\mathcal E|}{|\nabla\mathcal E|}.$$

These contractions are independent of coordinates and rotations of the chosen spatial triad; they still depend on the body’s four-velocity, since different observers split Weyl curvature into different electric and magnetic parts. The derivative in $\mathcal L_{\nabla\mathcal E}$ is spatially projected and its norm is measured in that same rest frame.

A local multipole expansion asks for

$$\epsilon_{\rm size} =\frac{R}{\min(\mathcal L_{\mathcal E}, \mathcal L_{\nabla\mathcal E})}\ll1.$$

That is not the same as weak material response. Compare the tidal acceleration across the body with its surface gravity:

$$\epsilon_{\rm tide} =\frac{R^3}{Gm} \sqrt{\frac{\mathcal E_{ab}\mathcal E^{ab}}{6}}.$$

Linear response further asks for a small deformation, whereas adiabatic elimination separately asks for a low forcing frequency:

$$\epsilon_Q =\frac{\sqrt{Q_{ab}Q^{ab}}}{mR^2}\ll1, \qquad \epsilon_\omega =\frac{|\omega_{\rm tide}|}{\omega_{\rm mode}}\ll1$$

The low-frequency condition can be relaxed by retaining the relevant mode explicitly; the small-deformation condition still applies to a linear mode model. These parameters diagnose different failures. A large mass ratio can produce $\epsilon_{\rm tide}\sim1$ while $R/r$ is still moderately small: the spatial multipole expansion may remain meaningful while perturbation about an undeformed equilibrium has already failed.

In Newtonian scaling, a companion of mass $M_c$ produces $|\mathcal E|\sim GM_c/r^3$. Equating tidal and self gravity gives

$$r_t \sim \xi_{\rm R}R \left(\frac{M_c}{m}\right)^{1/3}.$$

The coefficient $\xi_{\rm R}$ is not universal. It depends on the density profile, spin, orbital state, and the operational definition of onset. In a black-hole–neutron-star binary, disruption outside prompt capture is favored when this scale exceeds the relevant strong-field plunge scale. Replacing that scale by the test-particle Kerr ISCO is a useful estimate at large mass ratio, not an invariant criterion at comparable masses.

For two material bodies, compare $r_t$ with contact, mode-resonance, and orbital-instability scales. For a black-hole companion, compare it with horizon and capture scales. In either case, a statement such as “disruption occurs at $r=r_t$” is incomplete unless it specifies whether onset means a surface cusp, first unbound fluid, core destruction, waveform cutoff, or nonzero remnant mass.

What a worldline expansion can and cannot extend

A worldline action organizes finite-size effects schematically as

$$S_{\rm wl} =-m\int d\tau +\sum_n C_n\int d\tau\,\mathcal O_n +S_{\rm modes}.$$

Higher curvature operators improve the spatial expansion when successive powers of $R/\mathcal L$ remain ordered. Explicit modes repair the derivative expansion near a known resonance while their amplitudes remain small. Nonlinear response operators, for example terms cubic in the tidal field or quadratic in an induced mode, can extend the description beyond linear Love response.

This extension is real but conditional. Fully relativistic matching now provides static quadratic Love coefficients and their leading waveform contribution; Pani et al. find corrections that can reach about ten percent of the linear tidal contribution in selected late-inspiral neutron-star models in Pani et al. 2026, §§ II–V and § VI. Earlier nonlinear mode calculations likewise show that mode coupling can alter late-inspiral response within a specified Newtonian stellar model; see Yu et al. 2023, §§ 2–4 and § 5.

Neither calculation claims control through mass shedding. If a response series has the form

$$\frac{Q}{mR^2} =a_1\epsilon_{\rm tide} +a_2\epsilon_{\rm tide}^2 +a_3\epsilon_{\rm tide}^3+\cdots,$$

then $\epsilon_{\rm tide}\sim1$ removes a generic suppression between successive terms. Large mode amplitude can do the same to a mode-coupling series. Adding one quadratic coefficient does not prove that the cubic remainder is small.

There is a more qualitative boundary. Once matter leaves the surface, the system contains streams with scales and histories not encoded by a fixed set of body multipoles. Shocks generate entropy; composition and neutrino transport can change; self-bound fragments or a disk may form. One can enlarge the effective description by adding new collective variables, but a single worldline with fixed coefficients no longer has automatic error control.

Evidence comparison through 2026-08-12

Evidence cutoff. The comparison below includes literature checked through 2026-08-12. Results first released after that date are outside this assessment.

Approach	What it resolves	Controlled strength	Principal limitation near disruption
Curvature and Roche-scale comparison	Body size, self gravity, external curvature, capture or contact scale	Fast, invariantly expressible warning parameters	Order-unity onset coefficient and no post-shedding evolution
Explicit modes and nonlinear operators	Resonances, memory, selected nonlinear susceptibilities	Systematic while amplitudes and successive terms remain ordered	A finite truncation need not converge when the tide is order unity
Calibrated remnant or cutoff fit	Interpolation across simulated masses, compactnesses, and spins	Efficient inside a stated calibration domain with residuals	Fit boundary is not a physical theorem; threshold behavior and extrapolation can fail
Relativistic hydrodynamics	Surface shedding, shocks, streams, capture, disk and ejecta	Direct matter evolution for the chosen equation of state and microphysics	Finite resolution, atmosphere treatment, sparse parameter coverage, and model-dependent microphysics

The most-used black-hole–neutron-star scale fit illustrates both the value and danger of an empirical threshold. Label the black hole as body 1 and define $q=M_{\rm BH}/M_{\rm NS}\geq1$, $\nu=q/(1+q)^2$, neutron-star compactness $C=GM_{\rm NS}/(R_{\rm NS}c^2)$, and Kerr ISCO radius $\widehat R_{\rm ISCO}=R_{\rm ISCO}/(GM_{\rm BH}/c^2)$. Let $\widehat M_{\rm rem}=M_{\rm rem}/M_{\rm NS}^{b}$ denote the baryon mass outside the black hole about $10\,\mathrm{ms}$ after merger, normalized by the initial neutron star’s baryonic mass. Foucart, Hinderer, and Nissanke proposed

$$\widehat M_{\rm rem} =\left[ \max\left( \alpha\frac{1-2C}{\nu^{1/3}} -\beta\widehat R_{\rm ISCO}\frac{C}{\nu} +\gamma,\ 0 \right) \right]^\delta .$$

The four coefficients were calibrated to 75 numerical-relativity simulations. The source sample spanned approximately $q\in[1,7]$, aligned black-hole spin $\chi_{\rm BH}\in[-0.5,0.97]$, and $C\in[0.13,0.182]$, while excluding neutron-star spin, precession, eccentricity, and magnetic effects. The construction, calibration, and reported errors are given in Foucart, Hinderer, and Nissanke 2018, Eqs. (1)–(7) and Discussion.

Newer simulations do not support a single simplistic verdict. Chen et al. found that normalized remnant and ejecta masses and a waveform cutoff approximately group by $(q,\chi_{\rm BH},C)$ across their tested neutron-star masses, while more compact cases leave room for improved remnant fits; see Chen et al. 2024, §§ II D, III B–III C, and IV. By contrast, near the predicted threshold at $q=3$, Martineau et al. found the fitted critical spin broadly consistent but a biased slope of remnant mass versus spin, with discrepancies beyond the assigned model error for the DD2 equation of state; see Martineau et al. 2026, § II A, Eqs. (1)–(5), and § IV. Both can be true: a fit may capture a coarse threshold while missing the small but astrophysically important mass just outside it.

Two credible positions and their limits

Controlled-extension position. A worldline description should be extended with the modes and nonlinear operators that become important, matched to isolated-body calculations and tested against simulations. The success of dynamical modes before resonance and the 2026 relativistic quadratic-response calculation show that “linear Love response has failed” does not imply “every reduced description has failed.” This position is strongest while $\epsilon_{\rm size}\ll1$, resolved amplitudes remain bounded, and successive additions reduce discrepancies.

Its limitation is predictive control. A model with one more mode or one more nonlinear coefficient can fit closer to merger without demonstrating that the omitted tower is smaller. It must publish an order-by-order or matched residual, not merely a visually improved waveform.

Resolved-matter position. Mass shedding, shock formation, and topology change mark a qualitative boundary at which relativistic hydrodynamics is required. This view is supported by simulations in which disk and ejecta masses depend on near-surface matter and by the near-threshold sensitivity just described. A finite set of moments cannot represent an extended tidal tail after the body ceases to be a compact object.

Its limitation is coverage. Hydrodynamic simulations are expensive and cannot densely sample every mass ratio, spin orientation, equation of state, eccentricity, and microphysical model. Reduced descriptions remain essential for interpolation and for identifying which simulations are most informative.

The positions are not mutually exclusive. The unresolved question is how close to first mass shedding an augmented reduced model retains quantitative error control. A sensible hybrid uses worldline and mode dynamics in their demonstrated overlap, then hands off to resolved matter before the truncation loses hierarchy.

Uncertainties and open questions

Equation-of-state dependence enters through more than compactness. Two stars with similar $M/R$ can have different density profiles, sound speeds, composition gradients, crusts, and maximum masses. Spin can change the stellar equilibrium as well as the orbital capture scale. Precession makes the aligned Kerr-ISCO proxy inadequate. Eccentricity changes the tide impulsively. Temperature, magnetic fields, viscosity, neutrino transport, numerical atmosphere, extraction time, and the definition of “remnant” affect post-disruption observables.

The breakdown boundary should therefore be reported as a region with a declared outcome and uncertainty, not a universal radius. A proposed criterion should be checked against:

$$R,\quad r_t,\quad r_{\rm contact\ or\ capture},\quad \epsilon_Q,\quad \epsilon_\omega,\quad \text{and a resolved hydrodynamic outcome}.$$

The following evidence would change the present assessment:

• Matched convergence trigger: a worldline-plus-mode calculation and a relativistic hydrodynamic calculation starting from the same equation of state and binary data show order-by-order convergence as modes and operators are added, with residuals smaller than numerical errors up to a reproducible mass-shedding indicator. That would move the reduced-model boundary inward.
• Earlier hydrodynamic breakdown: cross-code, resolution-converged simulations with controlled equations of state could show that nominally higher operators cease to improve the same observables, or that shocks and surface outflow begin while the reduced model still predicts a small remainder. That would require switching to a hydrodynamic description at a larger separation.
• Threshold-calibration trigger: a dense, equation-of-state-controlled simulation campaign changes the critical surface or remnant-mass slope beyond the published fit and simulation uncertainties. The empirical fit must then be recalibrated rather than extrapolated.
• Invariant-diagnostic trigger: a curvature- and matter-based scalar predicts the same operational onset across mass ratio, spin, and equation of state with residuals below existing numerical error. It could replace the present collection of scale proxies.

Until one of these triggers is met, disruption fits should be used only inside their stated domains, nonlinear worldline extensions should carry truncation tests, and actual mass shedding or post-merger debris should be assigned to resolved matter.

Exercises

1. Roche scaling. Derive the mass-ratio dependence of $r_t$ from acceleration balance.

Solution

The differential acceleration across the body is of order

$$a_{\rm tide}\sim\frac{GM_cR}{r^3},$$

while surface self gravity is $a_{\rm self}\sim Gm/R^2$. Equating them gives

$$\frac{GM_cR}{r_t^3} \sim\frac{Gm}{R^2} \quad\Longrightarrow\quad r_t\sim R\left(\frac{M_c}{m}\right)^{1/3}.$$

The calculation fixes the scaling but not $\xi_{\rm R}$, because it ignores the density profile, rotation, relativity, and the definition of onset.

2. Scale comparison for a nonspinning black hole. Take $q=5$ and neutron-star compactness $C=0.17$. Compare the unit-coefficient Newtonian $r_t$ with the test-particle Schwarzschild ISCO.

Solution

The tidal estimate is

$$\frac{r_t}{R}=q^{1/3}=5^{1/3}\simeq1.71.$$

For a Schwarzschild hole, $R_{\rm ISCO}=6GM_{\rm BH}/c^2$, so

$$\frac{R_{\rm ISCO}}{R} =6qC=6(5)(0.17)=5.10.$$

The scale estimate places disruption well inside the test-particle ISCO and therefore suggests plunge before disruption. It is not a prediction: an order-unity Roche coefficient, comparable-mass corrections, and black-hole spin all shift the comparison. The exercise shows why both scales and their uncertainty must be stated.

3. Truncation warning. Suppose $a_1,a_2,a_3$ are all order unity. Compare the generic ratios of successive response terms at $\epsilon_{\rm tide}=0.2$ and $0.8$.

Solution

Ignoring accidental coefficient hierarchies, each successive-to-previous ratio is of order $\epsilon_{\rm tide}$. At $0.2$, the quadratic term is roughly twenty percent of the linear term and the cubic term roughly twenty percent of the quadratic term; a controlled improvement is plausible. At $0.8$, each omitted term can be comparable to the retained one. Computing only $a_2$ then does not establish a small remainder.

References

• Chen, Shichuan, Luohan Wang, Kota Hayashi, Kyohei Kawaguchi, Kenta Kiuchi, and Masaru Shibata. “Black Hole–Neutron Star Mergers with Massive Neutron Stars in Numerical Relativity.” Physical Review D 110, 063016 (2024). DOI; Open PDF.
• Foucart, Francois, Tanja Hinderer, and Samaya Nissanke. “Remnant Baryon Mass outside of the Black Hole after a Neutron Star–Black Hole Merger.” Physical Review D 98, 081501(R) (2018). DOI; Open PDF.
• Martineau, Tia, Francois Foucart, Mark A. Scheel, Matthew D. Duez, Lawrence E. Kidder, and Harald P. Pfeiffer. “Black Hole–Neutron Star Binaries near Neutron Star Disruption Limit in the Mass Regime of Event GW230529.” Classical and Quantum Gravity 43, 015015 (2026). DOI; Open PDF.
• Pani, Paolo, Massimiliano Maria Riva, Luca Santoni, Nikola Savić, and Filippo Vernizzi. “Nonlinear Relativistic Tidal Response of Neutron Stars.” Journal of High Energy Physics 2026, 074 (2026). DOI; Open PDF.
• Yu, Hang, Nevin N. Weinberg, Phil Arras, James Kwon, and Tejaswi Venumadhav. “Beyond the Linear Tide: Impact of the Non-Linear Tidal Response of Neutron Stars on Gravitational Waveforms from Binary Inspirals.” Monthly Notices of the Royal Astronomical Society 519, 4325–4343 (2023). DOI; Open article.