Relativistic Two-Body Dynamics

General relativity turns the Newtonian two-body problem into a meeting place of nonlinear field equations, multiple scales, radiation, gauge freedom, and compact-object structure. No single approximation covers every mass ratio, velocity, separation, spin, or encounter. This volume develops the complementary descriptions that do work—post-Newtonian and post-Minkowskian expansions, worldline effective theory, self-force theory, effective-one-body dynamics, and controlled numerical input—and teaches the invariant comparisons that make their overlap scientifically useful.

Helpful background. Begin with The Einstein Equation for the nonlinear field law and Weak Metric Perturbations and Gauge Freedom for controlled expansions and gauge dependence. Presymplectic Potentials and Currents becomes useful for the later treatment of conserved generators and Hamiltonian comparisons; the introductory route below starts from Newtonian orbital mechanics.

One problem, several controlled descriptions

If this is your first encounter with relativistic binary dynamics, begin with the familiar Newtonian reduction. Two bodies attract each other, but their relative separation obeys a one-particle problem with reduced mass $\mu=m_1m_2/(m_1+m_2)$ in the potential $-Gm_1m_2/r$. A circular orbit then has $\Omega^2=GM/r^3$ and binding energy $E_b=-GM\mu/(2r)$. Eliminating the separation gives $E_b/\mu=-(GM\Omega)^{2/3}/2$. That short calculation already contains the volume’s main strategy: predict a relation between measurable quantities and use a simple limit to test a harder calculation. The Newtonian two-body reduction provides the preparatory orbital mechanics.

General relativity changes each part of this picture. The gravitational field itself carries energy; radiation slowly changes the orbit; spin and internal structure add degrees of freedom; and the choice of coordinates changes intermediate formulas. “Post-Newtonian” (PN) means expanding beyond the slow-motion limit. “Post-Minkowskian” (PM) means expanding in gravitational coupling at fixed velocity. “Self-force” describes the correction caused by the small body’s own field. “Effective one body” (EOB) maps the interacting system into a useful effective description. These terms identify different tools for the same problem, with distinct accuracy limits.

Let the bodies have masses $m_1$ and $m_2$, total mass $M=m_1+m_2$, reduced mass $\mu=m_1m_2/M$, and symmetric mass ratio

$$\nu=\frac{\mu}{M}=\frac{m_1m_2}{(m_1+m_2)^2}, \qquad 0\lt \nu\leq\frac14.$$

For a quasi-circular orbit, the invariant frequency parameter

$$x=(GM\Omega)^{2/3}$$

reduces at Newtonian order to both $GM/r$ and $v^2$. Those equalities cease to be exact once $r$ and $v$ acquire coordinate dependence, but $x$ remains a useful label for circular-orbit observables. The pair $(\nu,x)$ already displays why one expansion is insufficient: post-Newtonian theory controls $x\ll1$ without requiring $\nu\ll1$, while gravitational self-force theory controls $\nu\ll1$ even in a strong field. Post-Minkowskian theory instead expands in $G$ or, for scattering at impact parameter $b$, in $GM/b$, without assuming a small asymptotic velocity.

Spin, eccentricity, tidal response, and radiation reaction add further independent ratios. A model is controlled only after it states which ratios are small, which quantities are held fixed, what boundary conditions select the solution, and how the error is expected to scale. The organizing lesson is therefore methodological: “PN,” “PM,” “self-force,” and “EOB” name approximation structures, not competing coordinate descriptions of one universal formula.

Modern post-Newtonian dynamics and its relation to Hamiltonian, multipolar, and radiative constructions are reviewed with explicit approximation order in Blanchet 2024, §§1 and 3, especially §§3.3–3.5. The small-mass-ratio construction, including the distinction between singular and regular fields and the need for long-time asymptotics, is developed in Barack and Pound 2019, §§3–5. These frameworks overlap, but their gauge-dependent intermediate fields need not agree term by term.

What counts as agreement

Two Hamiltonians can differ by a canonical transformation. Two equations of motion can differ by a coordinate change or by order reduction. A local self-force changes under a gauge transformation. Agreement must therefore be tested with a common observable and common physical data.

Depending on the regime, useful comparison quantities include:

• the binding energy $E(\Omega)$ for circular motion;
• periastron advance or ratios of fundamental frequencies for bound eccentric motion;
• Detweiler’s redshift invariant and spin-precession invariants at small mass ratio;
• the scattering angle $\chi(E,J)$, impulse, and radiated four-momentum for unbound motion;
• action variables and the radial action, which connect bound and scattering descriptions;
• asymptotic fluxes through null infinity and horizons, with a fixed normalization of time;
• tidal phase shifts and response coefficients after the object, matching convention, and frequency regime are specified.

An invariant comparison is still not automatic. The same masses, spins, energy, angular momentum, frame, asymptotic state, and perturbative truncation must be used. Dissipative observables require retarded boundary conditions; a time-symmetric conservative Hamiltonian cannot be compared directly with a fully retarded trajectory. A disagreement that vanishes after a canonical transformation is representational. A disagreement in the scattering angle at the same asymptotic energy and angular momentum is physical or signals an inconsistent truncation.

This emphasis follows the effective-one-body strategy: reorganize available information into a representation whose test-mass and weak-field limits are explicit, then compare invariant predictions rather than potential coefficients in isolation. A concise derivation of the energy map and its relation to the two-body problem appears in Damour 2014, §§4–5.

Find your route

For a first pass, read dimensionless parameters, then energy and frequency invariants, the worked 1PN circular-energy calculation, and flux balance. Try their solved exercises before choosing a specialist route below. You do not need amplitude methods, Kerr perturbations, or all thirteen chapters to understand the first relativistic corrections to a binary orbit.

Building analytic compact-binary dynamics

Begin with Scales, Invariants, and Regimes of the Two-Body Problem. Continue through Post-Newtonian Dynamics and Post-Minkowskian Dynamics and Scattering. These chapters establish the two principal weak-field organizations and the invariant quantities that survive a change of gauge or canonical variables.

Then choose a calculational language. Classical Amplitudes and the Gravity Two-Body Problem explains how on-shell quantum-field-theory tools yield classical impulses and scattering information after iterations and infrared effects are separated. Worldline Effective Field Theory organizes point-particle, spin, finite-size, dissipative, and scale-dependent operators through matching and power counting Porto 2016, §§6–7.

Extreme mass ratios and long inspirals

Read Gravitational Self-Force and Regularization before Extreme-Mass-Ratio Inspirals. The first chapter constructs motion from a singular perturbative field without treating the compact body as a literal distribution at every step. The second combines Kerr geodesic frequencies, flux balance, local forcing, transient resonances, and the transition to plunge into a long-time evolution.

This route distinguishes three expansions that are often conflated: expansion of the field in mass ratio, expansion of the orbital evolution in the ratio of orbital to radiation-reaction timescales, and expansion of a phase model to a stated absolute error. Passing a local field-equation check does not by itself control the accumulated phase over $O(\nu^{-1})$ orbital cycles.

Comparable masses, spin, and material structure

Effective-One-Body Dynamics assembles weak-field, test-mass, self-force, scattering, and numerical information into a resummed effective dynamics. Spin, Precession, and Multipolar Structure separates the physical spin tensor and multipoles from the representative worldline selected by a spin supplementary condition. Tides and Finite-Size Effects inserts black-hole or material response into orbital dynamics while preserving the boundary with the volumes that compute those responses.

The point is not to make every object look like a black hole. Spin-induced multipoles, adiabatic deformabilities, dynamical modes, absorption, and disruption carry distinct information. A Wilson coefficient is meaningful only with its operator normalization and matching prescription; a tidal deformability is meaningful only with its object, theory, multipole, frequency regime, and subtraction convention.

Dissipation, encounters, and model validation

Radiation Reaction and Multiscale Evolution connects balance laws, local dissipative forces, hereditary tails, averaging, resonances, and secular error. Capture, Scattering, and Strong-Field Encounters studies unbound and near-critical motion, where asymptotic energy and angular momentum replace circular frequency as the natural input.

Finish with Cross-Regime Matching, Calibration, and Consistency. This chapter turns overlaps into tests: PN coefficients must emerge from the low-velocity expansion of PM results; self-force information must enter finite-$\nu$ models through an explicitly chosen map; and numerical calibration must be separated from out-of-sample validation. A small residual on calibration data is not an uncertainty estimate.

Try a calculation

Each of these examples includes equations, downloadable code or data, and checks that you can repeat. Start with the first; the scalar wave solvers require more numerical preparation.

• Schwarzschild scattering and zoom–whirl motion connects a radial potential, an orbit and the logarithmic growth of the winding angle near capture.
• Mode-sum regularization computes a physical scalar force and shows why analytic subtraction improves convergence. The effective-source calculation reaches the same force component by evolving a residual wave field. Their comparison separates angular tails from wave-solution error.
• Released numerical-relativity scattering trajectories let you extract angles at three resolutions and see why a small resolution change can coexist with a larger extrapolation uncertainty.
• A failed physical calibration shows why good performance on a narrow withheld sample cannot rescue a model that already fails across its training domain.

Conventions that matter in two-body calculations

The site uses metric signature $(-,+,+,+)$, retains $G$ explicitly by default, and uses time dependence $e^{-i\omega t}$. Sources often set $G=c=1$ and may use the opposite metric signature. Results are translated as complete systems: mass dimensions, spin definitions, Green-function signs, Fourier phases, and boundary conditions move together.

Several local choices must always be stated:

• whether the mass ratio is $m_1/m_2$, $m_2/m_1$, $\nu$, or a perturbative parameter;
• whether $M$ is a sum of isolated masses, an ADM mass, or another energy;
• whether $a$ denotes Kerr spin length, dimensionless spin, or semimajor axis;
• which spin supplementary condition and representative worldline are used;
• whether a Hamiltonian is local, order-reduced, or contains a hereditary term;
• whether an observable is conservative, dissipative, time-symmetric, retarded, or orbit averaged;
• which frequencies use asymptotic time and which radial coordinate or tetrad defines a local quantity.

Throughout the volume, circular motion is a special case, not the definition of the two-body problem. Eccentric, inclined, precessing, and unbound configurations are introduced with their own invariant data. Likewise, a finite-order asymptotic series is not assigned an error merely by naming the next formal power; coefficient growth, logarithms, resonances, and proximity to a separatrix can change the useful remainder.

Scientific boundaries

This volume develops analytic and semi-analytic interacting-body dynamics. A complete binary model also needs results from several related subjects.

Black-Hole Perturbation Theory develops fixed-background black-hole response, master equations, boundary conditions, quasinormal spectra, and Love-number calculations. Here those results enter a binary as force, flux, response coefficient, or matching datum. Gravitational Radiation and Waveforms develops source-independent radiation theory, radiative multipoles, asymptotic waveform conventions, and waveform accuracy. Here radiation enters the orbital evolution and its evolution across several time scales.

Relativistic Matter and Compact Objects develops material equations of state, stellar response, and internal oscillation spectra. This volume asks how those properties alter two-body motion. Computational Relativity develops production spacetime evolution, numerical initial data, wave extraction, convergence, and verification. Here bounded numerical examples test analytic methods, compare invariant predictions, or examine calibration, with their extraction and error budgets visible. A coordinate-component benchmark, such as the scalar radial force, is identified explicitly and compared only after matching coordinates and normalization. Detector likelihoods, selection effects, and population inference belong to Gravitational-Wave Astronomy.

These boundaries prevent a common mistake: accuracy against one numerical waveform does not prove that an analytic model is controlled across parameter space, and an accurate orbital Hamiltonian does not by itself supply a detector-ready waveform.

Check your preparation

1. Two expansions. Suppose a circular binary has $\nu=10^{-5}$ and $x=0.2$. Which principal approximations are suggested by these numbers, and which one is not justified by smallness alone?

Solution

The very small $\nu$ supports a small-mass-ratio or self-force expansion, including strong-field motion on the background at leading order. The value $x=0.2$ is not parametrically small, so a straightforward low-order PN truncation is not justified merely by power counting. A PM treatment also needs scattering or weak-field data such as $GM/b$ and cannot be selected from $(\nu,x)$ alone for a bound orbit. An EOB representation may combine information from these limits, but its accuracy still requires independent validation.

Small mass ratio does not establish orbital stability. In the nonspinning Schwarzschild limit, $x=0.2$ means $r=5GM$, inside the innermost stable circular orbit at $6GM$. This circular orbit is unstable; a slowly evolving circular inspiral cannot be assumed there. Spin and the appropriate stability test must be specified before interpreting the parameters as an inspiral.

2. Gauge comparison. Two groups publish different coordinate accelerations at the same nominal PN order but agree on $E(\Omega)$. Is there necessarily a physical contradiction?

Solution

No. Coordinate accelerations can differ under a coordinate transformation, contact transformation, canonical transformation, or consistent order reduction. Agreement of the invariant circular binding-energy relation is evidence that the conservative physics may be equivalent. One should also compare another invariant and verify that masses, spins, truncation, and boundary conditions match before declaring equivalence.

3. Long-time accuracy. A local force is known with absolute error $O(\nu^3)$. Why does this statement alone not fix the accumulated inspiral phase error?

Solution

The phase is integrated over a radiation-reaction time containing $O(\nu^{-1})$ orbital cycles. Secular integration can promote a small local error, and oscillatory, averaged, resonant, and initial-data pieces scale differently. The phase claim therefore needs a multiscale expansion, the duration and observable, and an error budget that propagates the omitted force terms through the evolution.

References

• Barack, Leor, and Adam Pound. “Self-Force and Radiation Reaction in General Relativity.” Reports on Progress in Physics 82 (2019): 016904. DOI.
• Blanchet, Luc. “Post-Newtonian Theory for Gravitational Waves.” Living Reviews in Relativity 27 (2024): 4. DOI.
• Damour, Thibault. “The General Relativistic Two Body Problem.” In General Relativity, Cosmology and Astrophysics, edited by Jiří Bičák and Tomáš Ledvinka, 111–145. Cham: Springer, 2014. DOI.
• Porto, Rafael A. “The Effective Field Theorist’s Approach to Gravitational Dynamics.” Physics Reports 633 (2016): 1–104. DOI.

Further reading

• Schäfer, Gerhard, and Piotr Jaranowski. “Hamiltonian Formulation of General Relativity and Post-Newtonian Dynamics of Compact Binaries.” Living Reviews in Relativity 27 (2024): 2. DOI.