Compact Binary Inspiral

A compact binary inspiral is the loudest thing the universe knows how to say in gravitational waves. Two compact bodies — neutron stars or black holes, dense enough that their finite size does not matter until the very last orbits — circle their common centre of mass. The mass distribution is aspherical and rotating, so the system carries a time-varying mass quadrupole moment, and by the quadrupole formula that moment radiates. The waves carry off energy and angular momentum, so the orbit cannot stand still: it tightens, the bodies fall closer, they orbit faster, and — because the luminosity climbs steeply with frequency — the whole process runs away into a rising "chirp" that ends in merger. What makes the inspiral so valuable is that its early and middle stages are weak-field and slow-motion. A Newtonian orbit plus the leading radiation reaction — the back-reaction demanded by energy balance — already predicts the separation, frequency, and waveform in closed form, and the post-Newtonian expansion in $v/c$ sharpens that skeleton to template precision. Remarkably, at leading order the entire phase evolution collapses onto a single mass parameter, the chirp mass $\mathcal{M}=(m_1 m_2)^{3/5}/(m_1+m_2)^{1/5}$ : measure how fast the frequency sweeps and you have weighed the binary without ever resolving which body is which. The inspiral is where nearly every weak-field gravitational-wave idea meets — linearized theory's wave equation, the transverse-traceless gauge, the quadrupole formula as source, and the Isaacson stress-energy as the flux that closes the energy-balance loop. Hulse and Taylor confirmed the mechanism to a fraction of a percent decades before a wave was caught directly; GW150914 then caught one.

Key takeaways

A compact binary inspiral is the slow, secular shrinking of the orbit of two compact objects as gravitational waves carry away energy and angular momentum, ending in merger.
The radiation is quadrupolar — no monopole or dipole gravitational radiation — so the rotating mass quadrupole is the antenna, radiating at twice the orbital frequency.
Energy balance turns the quadrupole luminosity into Peters' decay law $\dot a\propto -a^{-3}$ , which integrates to a finite coalescence time $t_{\rm coal}\propto a_0^4$ .
The observable is a chirp, $\dot f=\frac{96}{5}\pi^{8/3}(G\mathcal{M}/c^3)^{5/3}f^{11/3}$ , so the leading waveform is fixed by one number, the chirp mass $\mathcal{M}$ .
The Hulse–Taylor pulsar's decay matched the prediction to better than a percent; GW150914 ( $\approx36$ and $29\,M_\odot$ ) was the first direct detection, GW170817 the first multi-messenger one.
Key idea: the orbit's time-varying mass quadrupole radiates at a rate fixed by the quadrupole formula; equating that luminosity to the loss of orbital binding energy gives a closed evolution equation whose only mass parameter at leading order is the chirp mass $\mathcal{M}=(m_1 m_2)^{3/5}/(m_1+m_2)^{1/5}$ .

The core insight

The orbit's time-varying mass quadrupole radiates gravitational waves at a rate fixed by the quadrupole formula; equating that luminosity to the loss of orbital binding energy gives a closed evolution equation whose only mass parameter at leading order is the chirp mass $\mathcal{M}=(m_1 m_2)^{3/5}/(m_1+m_2)^{1/5}$ .

Where this lives today

The inspiral is the working core of gravitational-wave astronomy. Because its waveform is known analytically — the Newtonian chirp refined to high post-Newtonian order — searches cross-correlate detector data against banks of hundreds of thousands of templates, and the chirp mass, controlling the bulk of the phase, is pinned down even for weak signals: GW150914 was found this way. Once $\mathcal{M}$ is known the strain amplitude fixes the luminosity distance, turning compact binaries into "standard sirens" for the Hubble constant, independent of the distance ladder, as GW170817 demonstrated. The coalescence-time formula tells which binaries merge within a Hubble time, and thus which formation channels — isolated evolution versus dynamical capture — supply the observed rate. For binary neutron stars the inspiral predicts the merger time and sky position precisely enough for telescopes to catch the counterpart: GW170817's chirp was followed by a gamma-ray burst and a kilonova, confirming that such mergers forge heavy elements and that gravitational and electromagnetic waves travel at the same speed to one part in $10^{15}$ . At the peak, two merging black holes briefly radiate $\sim10^{-2}\,c^5/G\approx3\times10^{49}\ \mathrm{W}$ — for a few milliseconds, more than all the stars in the observable universe. The frontiers now are LISA, which will hear massive-black-hole and extreme-mass-ratio inspirals for months, and pulsar timing arrays already sensing the hum of supermassive binaries.

The physical picture

Picture the two stars as a slowly turning dumbbell. Its mass distribution is aspherical and rotating, so the quadrupole moment oscillates — and because a quadrupole looks the same after half a turn, it oscillates twice per orbit. That oscillating quadrupole is the antenna: waves stream out at twice the orbital frequency, paid for out of the binding energy. The first figure fixes the mechanism and the two relations that drive everything.

Figure

A binary as a rotating mass quadrupole: two bodies m₁ and m₂ separated by a orbit at angular frequency Ω set by Kepler Ω² = GM/a³ and radiate at f_GW = 2 f_orb with luminosity L_GW ∝ Ω⁶ — the steep scaling that makes the inspiral run away

Because the binding energy grows more negative as the stars approach, losing energy tightens the orbit and speeds it up — the opposite of everyday friction. A bound orbit has, in this sense, a negative heat capacity: remove energy and it gets "hotter" (faster). Closer bodies orbit faster, radiate more strongly, and lose energy faster still, so the inspiral accelerates without limit: frequency and amplitude both rise into the chirp that ends in merger. The second figure is what a detector records.

Figure

The chirp waveform h(t): as the orbit tightens, frequency and amplitude both climb and end abruptly at merger, with f(t) ∝ (t_c − t)^(−3/8). GW150914 swept from about 35 Hz to 250 Hz in roughly 0.2 s over some 10 cycles; a 1.4 + 1.4 M⊙ neutron-star binary sweeps from 10 Hz over about 16000 cycles in about 1000 s

The radiation starts at the quadrupole — not, as a radiating charge might suggest, at the dipole — by pure conservation law: the mass monopole (total mass) and the mass dipole's derivative (total momentum) are both conserved, so the lowest time-varying multipole is the quadrupole. This is why gravitational radiation is so much feebler than electromagnetic dipole radiation, and why only the most violent, compact systems become detectable.

The core equations

Everything begins with the radiation field of a time-varying quadrupole. Projected into the transverse-traceless gauge at the retarded time, the strain is the second time derivative of the source's mass quadrupole, weakened by $1/r$ and the small coupling $G/c^4$ .

\bar{h}_{ij}^{\rm TT}(t,\mathbf{x}) = \frac{2G}{c^4 r}\,\ddot{I}_{ij}^{\rm TT}\!\left(t - r/c\right) \qquad (1)

where: $\bar{h}_{ij}^{\rm TT}$ — transverse-traceless metric perturbation (the observable strain); $I_{ij}$ — mass quadrupole moment of the source; $r$ — distance to the observer; $t-r/c$ — retarded time; $G,c$ — Newton's constant and the speed of light. Reading it: the strain measures how fast the source's shape changes, and the tiny prefactor $2G/c^4$ is why astrophysical quadrupoles produce strains near $10^{-21}$ . The power carried off is the Isaacson-averaged flux, which for the quadrupole reduces to the square of the third time derivative of the trace-free moment, averaged over several cycles.

L_{\rm GW} = -\frac{\mathrm{d}E}{\mathrm{d}t} = \frac{G}{5c^5}\,\big\langle \dddot{\mathcal{I}}_{ij}\,\dddot{\mathcal{I}}_{ij} \big\rangle \qquad (2)

where: $L_{\rm GW}$ — power radiated as gravitational waves; $\mathcal{I}_{ij}$ — reduced (trace-free) mass quadrupole moment; $\langle\cdot\rangle$ — average over several orbits. Reading it: three time derivatives make the luminosity ferociously sensitive to how fast the quadrupole varies — every factor of frequency amplified sixfold once squared. Specialize to a circular binary and that sensitivity becomes the characteristic $\Omega^6$ law — the single most important scaling in the subject: doubling the orbital frequency raises the luminosity sixty-fourfold.

L_{\rm GW} = \frac{32}{5}\frac{G}{c^5}\,\mu^2 a^4 \Omega^6 = \frac{32}{5}\frac{G^4}{c^5}\frac{m_1^2 m_2^2\, M}{a^5} \qquad (3)

where: $\mu=m_1 m_2/(m_1+m_2)$ — reduced mass; $M=m_1+m_2$ — total mass; $a$ — orbital separation; $\Omega$ — orbital angular frequency. Reading it: the second form trades $\Omega$ for $a$ via Kepler, exposing the brutal $a^{-5}$ growth as the orbit tightens. The dynamics feeding this luminosity are Newtonian: a Keplerian angular frequency and a negative binding energy that is the fuel tank for the radiation.

\Omega^2 = \frac{G M}{a^3},\qquad E = -\frac{G m_1 m_2}{2a} \qquad (4)

where: $M=m_1+m_2$ — total mass; $E$ — orbital binding energy, negative because bound. Reading it: since $E<0$ and $|E|\propto1/a$ , draining energy forces $a$ down — the orbit tightens rather than widens, the counterintuitive heart of the inspiral. Equating $\dot E$ to $-L_{\rm GW}$ gives Peters' orbital-decay law, and integrating it gives a finite time to merge.

\frac{\mathrm{d}a}{\mathrm{d}t} = -\frac{64}{5}\frac{G^3 m_1 m_2 M}{c^5 a^3},\qquad t_{\rm coal} = \frac{5}{256}\frac{c^5\, a_0^4}{G^3 m_1 m_2 M} \qquad (5)

where: $\mathrm{d}a/\mathrm{d}t$ — rate of shrinkage; $t_{\rm coal}$ — time to merge from initial separation $a_0$ . Reading it: $\dot a\propto a^{-3}$ integrates to $a(t)^4=a_0^4-(256/5)(G^3 m_1 m_2 M/c^5)\,t$ , so the separation reaches zero in finite time — the orbit does not decay asymptotically, it coalesces. At leading order the two masses enter the waveform only through one combination, the chirp mass.

\mathcal{M} = \frac{(m_1 m_2)^{3/5}}{(m_1+m_2)^{1/5}} = \mu^{3/5} M^{2/5} \qquad (6)

where: $\mathcal{M}$ — the single mass parameter governing the leading inspiral. Reading it: $\mathcal{M}$ is a weighted mean of the masses, always $\le M/4^{1/5}$ (equal to $M/2^{6/5}$ for equal masses) — what a detector measures first and most precisely. Writing the decay in terms of the gravitational-wave frequency $f=2f_{\rm orb}=\Omega/\pi$ gives the chirp equation, the phase law matched-filter searches ride.

\dot{f} = \frac{96}{5}\pi^{8/3}\left(\frac{G\mathcal{M}}{c^3}\right)^{5/3} f^{11/3} \qquad (7)

where: $f$ — gravitational-wave frequency, twice the orbital; $\dot f$ — the "chirp rate" measured to extract $\mathcal{M}$ . Reading it: integrating gives $f(t)\propto(t_{\rm coal}-t)^{-3/8}$ , the runaway sweep, and the only mass information in the leading phase is $\mathcal{M}$ . The strain amplitude, read off with the phase, scales as $h\simeq\dfrac{4}{r}\big(\dfrac{G\mathcal{M}}{c^2}\big)^{5/3}\big(\dfrac{\pi f}{c}\big)^{2/3}\cos\Phi(t)$ : no independent mass appears there either, so once $\mathcal{M}$ is fixed by the phase the amplitude measures the distance $r$ — the "standard siren".

Derivation AFrom the quadrupole formula to the circular-binary luminosity

The goal is to turn the abstract quadrupole luminosity, equation (2), into the concrete $\Omega^6$ law of equation (3) for a circular binary, so the steep frequency dependence — the reason the inspiral accelerates — emerges from an explicit calculation rather than a scaling argument. We build the reduced quadrupole of two point masses on a circular orbit, take three time derivatives, square and orbit-average, and track the coefficient to the standard $32/5$ .

Step 1. Place the relative coordinate on a circle, $x=a\cos\Omega t$ , $y=a\sin\Omega t$ , $z=0$ , and form the reduced quadrupole $\mathcal{I}_{ij}=\mu(x_i x_j-\tfrac13\delta_{ij}a^2)$ . Why: a two-body problem with reduced mass $\mu$ and separation $a$ has exactly the quadrupole of a single mass $\mu$ tracing the relative orbit, so one calculation covers any mass ratio.

Step 2. Use double-angle identities, $x^2=\tfrac{a^2}{2}(1+\cos2\Omega t)$ , $xy=\tfrac{a^2}{2}\sin2\Omega t$ , $y^2=\tfrac{a^2}{2}(1-\cos2\Omega t)$ , so every component of $\mathcal{I}_{ij}$ oscillates at $2\Omega$ . Why: a quadrupole is invariant under a half-turn, so it repeats twice per orbit — the origin of $f_{\rm GW}=2f_{\rm orb}$ .

Step 3. Differentiate three times; each derivative brings a factor $2\Omega$ , giving $\dddot{\mathcal{I}}_{xx}=4\mu a^2\Omega^3\sin2\Omega t$ , $\dddot{\mathcal{I}}_{xy}=-4\mu a^2\Omega^3\cos2\Omega t$ , $\dddot{\mathcal{I}}_{yy}=-\dddot{\mathcal{I}}_{xx}$ . Why: the three derivatives convert the modest amplitude $\mu a^2$ into $\mu a^2\Omega^3$ , and squaring hands us $\Omega^6$ .

Step 4. Square, sum, and orbit-average with $\langle\sin^2\rangle=\langle\cos^2\rangle=\tfrac12$ : $\langle\dddot{\mathcal{I}}_{ij}\dddot{\mathcal{I}}_{ij}\rangle=\langle\dddot{\mathcal{I}}_{xx}^2+2\dddot{\mathcal{I}}_{xy}^2+\dddot{\mathcal{I}}_{yy}^2\rangle=32\,\mu^2 a^4\Omega^6$ . Why: this bookkeeping fixes the exact coefficient, not a mere proportionality.

Step 5. Multiply by $G/5c^5$ to get $L_{\rm GW}=\tfrac{32}{5}\tfrac{G}{c^5}\mu^2 a^4\Omega^6$ , then use $\Omega^2=GM/a^3$ and $\mu=m_1m_2/M$ to reach $L_{\rm GW}=\tfrac{32}{5}\tfrac{G^4}{c^5}m_1^2m_2^2 M/a^5$ . Why: the first form gives the frequency law that drives the chirp, the second the separation law $L\propto a^{-5}$ needed for energy balance. *Verified: coefficient $32/5$ and $L_{\rm GW}=\tfrac{32}{5}\tfrac{G^4}{c^5}m_1^2m_2^2M/a^5$ reproduced against Peters & Mathews (1963), Maggiore Vol. 1 §3.4, and MTW §36.*

Go deeper · Derivation B · 1 minEnergy balance and the closed-form chirp

Derivation BEnergy balance and the closed-form chirp

The goal is to derive the chirp equation (7), $\dot f\propto f^{11/3}$ , and show that all the two-body mass dependence collapses into the single chirp mass $\mathcal{M}$ — the result that makes the inspiral a precision mass-measuring instrument. We treat an adiabatically shrinking quasi-circular orbit, balance radiated power against binding energy to get Peters' decay, then re-express everything in the observable frequency.

Step 1. Differentiate $E=-Gm_1m_2/2a$ to get $\dot E=(Gm_1m_2/2a^2)\dot a$ , and impose energy balance $\dot E=-L_{\rm GW}$ with the circular luminosity (3). Why: adiabaticity is the crucial assumption — over one orbit the geometry is Keplerian, and the only slow change is the secular drain of $E$ into radiation.

Step 2. Equate $(Gm_1m_2/2a^2)\dot a=-\tfrac{32}{5}(G^4/c^5)m_1^2m_2^2M/a^5$ and solve: $\dot a=-\tfrac{64}{5}(G^3/c^5)m_1m_2M/a^3$ (Peters' law). Integrating gives $a^4=a_0^4-\tfrac{256}{5}(G^3m_1m_2M/c^5)\,t$ , hence $t_{\rm coal}=\tfrac{5}{256}\,c^5 a_0^4/(G^3m_1m_2M)$ . Why: the $a^{-3}$ law makes the shrinkage a runaway — the closer the orbit, the faster it closes — and fixes a finite coalescence time.

Step 3. Convert separation to frequency. Kepler's law gives $f=\Omega/\pi=\pi^{-1}\sqrt{GM/a^3}$ , so $f\propto a^{-3/2}$ , $\dot f=-\tfrac32(f/a)\dot a$ , and $a=(GM/\pi^2 f^2)^{1/3}$ . Why: the detector sees $f$ , not $a$ , so the physics must be re-expressed in the observable — clean here because both relations are pure power laws.

Step 4. Substitute Peters' $\dot a$ and the inverted $a$ into $\dot f=-\tfrac32(f/a)\dot a=\tfrac{96}{5}(fG^3\mu M^2/c^5 a^4)$ with $a^4=(GM)^{4/3}\pi^{-8/3}f^{-8/3}$ ; the mass factors collapse to $\mu M^{2/3}=(\mu^{3/5}M^{2/5})^{5/3}=\mathcal{M}^{5/3}$ . Why: this is where the chirp mass appears — the powers $3/5$ and $2/5$ are exactly those that make $\mu M^{2/3}$ a perfect $5/3$ power, so no other mass combination survives at this order.

Step 5. Assemble: $\dot f=\tfrac{96}{5}\pi^{8/3}(G\mathcal{M}/c^3)^{5/3}f^{11/3}$ . Integrating from $f$ to the divergence gives $f\propto(t_{\rm coal}-t)^{-3/8}$ . Why: the $-3/8$ power is the analytic signature of the chirp, and its measured shape independently pins down $\mathcal{M}$ , over-determining the fit and testing general relativity. *Verified: coefficient $96/5$ , the exponent $11/3$ , the chirp-mass collapse $\mu M^{2/3}=\mathcal{M}^{5/3}$ , and the $-3/8$ power law reproduced against Peters (1964), Maggiore Vol. 1 §4.1, and Carroll §7.6. Numerically $GM_\odot/c^3=4.93\times10^{-6}\ \mathrm{s}$ .*

Go deeper · Derivation C · 1 minThe eccentric enhancement and the Hulse–Taylor orbital decay

Derivation CThe eccentric enhancement and the Hulse–Taylor orbital decay

The goal is to extend energy balance to an eccentric orbit and predict the period derivative $\dot P_b$ that pulsar timing measures — the quantity that first confirmed gravitational radiation quantitatively, decades before a direct detection. Eccentric orbits radiate far more strongly than circular ones of the same period, because the bodies whip through periastron where the $a^{-5}$ luminosity spikes; we capture that with the Peters–Mathews enhancement factor and convert the energy loss into a shrinking period.

Step 1. Orbit-average the luminosity over a Keplerian ellipse: the circular result is multiplied by $f(e)=\big(1+\tfrac{73}{24}e^2+\tfrac{37}{96}e^4\big)/(1-e^2)^{7/2}$ , giving $\langle\dot E\rangle=-\tfrac{32}{5}(G^4/c^5)(m_1^2m_2^2M/a^5)f(e)$ . Why: the factor blows up as $e\to1$ , encoding the radiation bursts at each periastron — for PSR B1913+16 with $e\approx0.62$ it is already about a factor of 12.

Step 2. Relate period to energy: Kepler's $P_b^2\propto a^3$ and $|E|\propto1/a$ give $\dot P_b/P_b=-\tfrac32\,\dot E/|E|$ , so lost binding energy shortens the period. Why: the same negative-heat-capacity logic as the circular case, now read out as a decreasing period that a pulsar clock measures superbly.

Step 3. Express in the measured period: $\dot P_b=-\tfrac{192\pi G^{5/3}}{5c^5}\big(\tfrac{P_b}{2\pi}\big)^{-5/3}\,\dfrac{m_1 m_2}{(m_1+m_2)^{1/3}}\,f(e)$ . Why: $P_b$ , $e$ , and the masses all come from pulse arrival times, so this is a parameter-free prediction — nothing is tuned to fit the decay.

Step 4. Insert the values for PSR B1913+16 ( $P_b\approx0.323\ \mathrm{day}$ , $e\approx0.617$ , $m_1,m_2\approx1.44,1.39\,M_\odot$ ) to get $\dot P_b^{\rm GR}\approx-2.40\times10^{-12}\ \mathrm{s/s}$ , matching the measured decay to better than a percent after a small galactic correction. Why: this was the first quantitative proof that gravitational waves carry energy as the quadrupole formula demands — the bedrock beneath every inspiral template. *Verified: coefficients $73/24$ , $37/96$ , prefactor $192\pi/5$ , and $\dot P_b^{\rm GR}\approx-2.40\times10^{-12}\ \mathrm{s/s}$ recomputed against Peters (1964) and the Weisberg–Taylor timing values.*

Limiting cases

The inspiral is described by different tools in different regimes, bounded by the orbital velocity $v/c$ and the separation relative to the innermost stable circular orbit.

Case	Result	Meaning
Newtonian + radiation reaction	$v/c\ll1,\ r\gg r_{\rm ISCO}$	Quasi-circular Keplerian orbit with adiabatic quadrupole energy loss; gives the closed-form chirp, the basis of leading-order templates.
Post-Newtonian inspiral	$v/c\lesssim0.3$	Expansion in $(v/c)^2$ adds spin, tail, and higher-multipole corrections; converges through the late inspiral but not at merger.
Late inspiral / merger	$r\sim r_{\rm ISCO}=6GM/c^2$	Near the innermost stable circular orbit the PN series ceases to converge; full numerical relativity is required.
Extreme mass ratio	$q=m_2/m_1\ll1$	A small body spiralling into a massive hole is treated by black-hole perturbation theory and the self-force, not the PN expansion.
Highly eccentric	$e\to1$	Radiation comes in periastron bursts; the factor $f(e)\sim(1-e^2)^{-7/2}$ diverges and the orbit circularizes as it decays.

Common trapAn orbiting mass radiates like an accelerating charge, so the leading gravitational radiation is dipole.

It is not — and the difference is a conservation law, not a technicality. For electromagnetism the dipole $\sum q_i\mathbf{r}_i$ has a nonzero second derivative, so accelerating charges radiate at dipole order. For gravity the "charge" is mass: the mass dipole $\sum m_i\mathbf{r}_i=M\mathbf{R}_{\rm cm}$ has first derivative equal to the total momentum, which is conserved, so $\ddot{\big(\text{dipole}\big)}=0$ and there is no mass-dipole radiation; the monopole (total mass) is conserved too. The lowest radiating moment is therefore the quadrupole, which is why gravitational radiation is intrinsically weak. A second, separate trap: the Newtonian-plus-radiation-reaction treatment holds only while $v\ll c$ and the bodies are well separated. Near merger the post-Newtonian series stops converging — do not extrapolate the closed-form chirp through $r_{\rm ISCO}$ ; that regime needs numerical relativity.

Worked examples

1 · The chirp mass of GW150914 (confidence builder)

GW150914 was two black holes of source-frame masses $m_1\approx36\,M_\odot$ and $m_2\approx29\,M_\odot$ . Find the chirp mass and compare it with the total mass.

Step 1. Form the products: $m_1 m_2=1044\,M_\odot^2$ , $M=65\,M_\odot$ , so $\mathcal{M}=(m_1 m_2)^{3/5}/M^{1/5}=1044^{0.6}/65^{0.2}$ .

Step 2. Evaluate: $1044^{0.6}=\exp(0.6\times6.951)=64.7$ and $65^{0.2}=\exp(0.2\times4.174)=2.305$ , so $\mathcal{M}=64.7/2.305=\mathbf{28.1\,M_\odot}$ (source frame).

Step 3. Interpret. The chirp mass sits well below the total mass $65\,M_\odot$ — for any binary $\mathcal{M}\le M/4^{1/5}\approx0.76\,M$ , here $\mathcal{M}\approx0.43\,M$ because the masses are unequal. What the detector actually measured is the redshifted $(1+z)\mathcal{M}\approx30\,M_\odot$ at $z\approx0.09$ ; correcting for redshift recovers the $28\,M_\odot$ intrinsic value. The chirp mass is $\mathcal{M}\approx28\,M_\odot$ (source frame), about $30\,M_\odot$ in the detector frame.

2 · How long a neutron-star inspiral lives in the detector band (fully worked)

Two $1.4\,M_\odot$ neutron stars enter a ground-based detector's band at $f=10\ \mathrm{Hz}$ . Find the chirp mass, the time to merger, and roughly how many wave cycles the detector records.

Step 1. Chirp mass. With $\mu=0.7\,M_\odot$ , $M=2.8\,M_\odot$ : $\mathcal{M}=\mu^{3/5}M^{2/5}=0.7^{0.6}\times2.8^{0.4}\,M_\odot=0.807\times1.510\,M_\odot=\mathbf{1.22\,M_\odot}$ , so $G\mathcal{M}/c^3=1.22\times4.93\times10^{-6}=6.01\times10^{-6}\ \mathrm{s}$ .

Step 2. Time to merger. Integrating the chirp equation gives $\tau=\tfrac{5}{256}(G\mathcal{M}/c^3)^{-5/3}(\pi f)^{-8/3}$ . With $(G\mathcal{M}/c^3)^{5/3}=(6.01\times10^{-6})^{5/3}=1.98\times10^{-9}$ and $(\pi f)^{-8/3}=(31.4)^{-8/3}=1.02\times10^{-4}$ , $\tau=0.01953\times(1/1.98\times10^{-9})\times1.02\times10^{-4}\approx\mathbf{1.0\times10^{3}\ s}\approx17\ \mathrm{min}$ .

Step 3. Number of cycles. From $f$ to merger, $N=\tfrac{3}{5k}f^{-5/3}$ with $k=\tfrac{96}{5}\pi^{8/3}(G\mathcal{M}/c^3)^{5/3}=8.1\times10^{-7}$ , giving $N\approx\mathbf{1.6\times10^{4}}$ cycles. By contrast GW150914's far larger chirp mass swept it from $\sim35\ \mathrm{Hz}$ to merger in $\sim0.2\ \mathrm{s}$ over about ten cycles.

Step 4. The lesson. A light binary lingers for thousands of slow cycles, so its phase — and hence $\mathcal{M}$ — is measured superbly; a heavy binary flashes through in a fraction of a second. Either way the sweep rate reads out the chirp mass, and the analytic picture holds until roughly the Schwarzschild ISCO frequency $f_{\rm ISCO}=c^3/(6^{3/2}\pi GM)\approx68\ \mathrm{Hz}$ for a $65\,M_\odot$ binary. $\mathcal{M}=1.22\,M_\odot$ , $\tau\approx1000\ \mathrm{s}$ , $N\approx16000$ cycles from 10 Hz.

Physics of the portrait

The compact binary inspiral was a theorist's construction long before it was an observation, but the number that made it undeniable came from a radio telescope — a single binary pulsar measured tightening its orbit exactly as the quadrupole formula demanded, the first hard evidence that gravitational waves carry energy.

Origins · 1 minHulse and Taylor and how this was discovered

Hulse and Taylor

Hulse b. 1950 · Taylor b. 1941 · the binary pulsar and the first evidence for gravitational radiation

In 1974 Russell Hulse, then a graduate student, and his advisor Joseph Taylor found PSR B1913+16 in an Arecibo survey — a pulsar whose period wobbled with a 7.75-hour rhythm from Doppler shift, betraying an unseen neutron-star companion. It was a laboratory of extraordinary quality: a clock of atomic precision in a strong field, so every relativistic subtlety (periastron advance forty thousand times Mercury's, gravitational redshift, the Shapiro delay) could be read from pulse arrival times. Most consequentially, general relativity predicted the orbit would lose energy to gravitational radiation and shrink, shortening the period by the tiny $\dot P_b\approx-2.4\times10^{-12}\ \mathrm{s/s}$ of Peters' formula. By 1982 the timing traced a parabola of cumulative shift agreeing to better than a percent — the first quantitative proof that gravitational waves exist, for which Hulse and Taylor shared the 1993 Nobel Prize. Their pulsar did not detect a wave; it detected the recoil of emitting one, validating the quadrupole-radiation edifice on which, three decades later, LIGO's templates were built.

History

1963

— Peters and Mathews compute gravitational-wave power from a Keplerian binary; Peters gives the decay laws the next year

1974

— Hulse and Taylor discover the binary pulsar PSR B1913+16 at Arecibo

1982

— the measured orbital decay of B1913+16 matches the quadrupole prediction to better than a percent

2015

— LIGO records GW150914, the first direct detection of a compact-binary inspiral and merger

2017

— GW170817, a binary neutron star, is seen in gravitational and electromagnetic waves together

Going further

Texts. Misner, Thorne & Wheeler, Gravitation, §36 · Wald, General Relativity, §4.4 · Carroll, Spacetime and Geometry, §7.6 · Hartle, Gravity, Ch. 16, 23 · Schutz, A First Course in General Relativity, §9.3 · Maggiore, Gravitational Waves Vol. 1, Ch. 3–4 · Peters & Mathews, Phys. Rev. 131, 435 (1963); Peters, Phys. Rev. 136, B1224 (1964).

Provenance. The quadrupole luminosity, the Peters decay and coalescence-time laws, and the eccentric enhancement follow Peters & Mathews (1963) and Peters (1964), with textbook treatments in Maggiore Vol. 1 §§3–4, Carroll §7.6, MTW §36, and Schutz §9.3. A verifier should recompute the $32/5$ luminosity coefficient and its $a^{-5}$ form; the $64/5$ in $\dot a$ and $5/256$ coalescence prefactor; the $96/5$ chirp coefficient, the $11/3$ exponent, and the $-3/8$ power law; the eccentricity coefficients $73/24$ , $37/96$ with prefactor $192\pi/5$ ; and the worked numbers $GM_\odot/c^3=4.93\times10^{-6}\ \mathrm{s}$ , $\mathcal{M}_{\rm GW150914}\approx28\,M_\odot$ , $\mathcal{M}_{\rm NS-NS}=1.22\,M_\odot$ ( $\tau\approx1000\ \mathrm{s}$ , $N\approx16000$ from 10 Hz), $\dot P_b^{\rm GR}\approx-2.40\times10^{-12}\ \mathrm{s/s}$ for PSR B1913+16 ( $e\approx0.617$ ), and $f_{\rm ISCO}\approx68\ \mathrm{Hz}$ for $M=65\,M_\odot$ . GW150914 ( $\approx36,29\,M_\odot$ ) and GW170817 are LIGO–Virgo values; ISCO at $6GM/c^2$ is the Schwarzschild fiducial.

Prerequisites. Quadrupole Formula · Gravitational-Wave Equation · Isaacson Stress-Energy of Gravitational Waves · Transverse-Traceless Gauge · Plus and Cross Polarizations. Leads to. Multipole Expansion of Gravitational Radiation · Geodesic Deviation and Tidal Action of Waves · Helicity and Spin-2 Nature of Gravitons.

Ask the tutor: Ask me to run the chirp equation on any pair of masses — we can compute the chirp mass, the time-to-merger from any starting frequency, and the number of cycles a detector would hear.

Why does the binary speed up as it loses energy — isn't that backwards?

It looks backwards only if you forget the orbit is bound, so its total energy is negative. The binding energy $E=-Gm_1m_2/2a$ becomes more negative as $a$ shrinks, so radiating energy away drives $a$ down — and by Kepler's law a smaller orbit has a higher frequency. A bound system behaves as if it had a negative heat capacity: take energy out and it gets "hotter", faster and more tightly bound — the same reason a satellite in a decaying orbit speeds up as drag lowers it. This is the engine of the inspiral.

What exactly is the chirp mass, and why does it dominate the signal?

The chirp mass $\mathcal{M}=(m_1 m_2)^{3/5}/(m_1+m_2)^{1/5}$ is the only mass parameter in the leading-order phase and amplitude. The reason is algebraic: pushing the quadrupole luminosity through energy balance and re-expressing the decay in the observed frequency, all the mass factors collapse into $\mu M^{2/3}=\mathcal{M}^{5/3}$ . Because a detector measures the accumulated phase over many cycles very precisely, $\mathcal{M}$ is typically the best-determined parameter of any inspiral, often to a fraction of a percent, even when the individual masses stay uncertain and only the higher post-Newtonian terms can separate them.

Why is there a factor of two between the orbital and gravitational-wave frequencies?

Because the source is a quadrupole, not a dipole. A quadrupole is unchanged under a half-turn — a dumbbell pointing left-right looks identical to one pointing right-left — so over one orbit the quadrupole pattern repeats twice, and the wave oscillates at $f_{\rm GW}=2f_{\rm orb}$ . Concretely, building the quadrupole from $\cos\Omega t$ and $\sin\Omega t$ , the double-angle identities turn everything into $\cos2\Omega t$ and $\sin2\Omega t$ . Worth remembering: a binary orbiting at 100 Hz radiates gravitational waves at 200 Hz.

How does one measurement of a chirp yield both the mass and the distance?

The two pieces of the waveform carry different information. The phase — how the frequency sweeps upward — depends only on $\mathcal{M}$ through $\dot f\propto\mathcal{M}^{5/3}f^{11/3}$ , so tracking the sweep measures the chirp mass with no reference to distance. The amplitude scales as $h\propto\mathcal{M}^{5/3}f^{2/3}/r$ , in which the only remaining unknown once $\mathcal{M}$ is fixed is the distance $r$ . Phase gives the mass, amplitude gives the distance — no external calibration needed, which is why compact binaries are "standard sirens". The catch: the amplitude also depends on orientation and sky position, so multi-detector localization sharpens the distance.

Where does the analytic inspiral stop being valid?

The closed-form chirp assumes weak fields and slow motion — the post-Newtonian expansion in $v/c$ . As the bodies approach the innermost stable circular orbit, near $r_{\rm ISCO}=6GM/c^2$ , the orbital speed climbs toward a fraction of $c$ and the expansion stops converging; beyond it there is no stable orbit, so the bodies plunge and merge. That strong-field regime needs full numerical relativity, with the remnant's ringdown handled by black-hole perturbation theory. Modern models stitch the analytic inspiral onto numerical merger simulations into one inspiral–merger–ringdown template — for GW150914 the inspiral ended near $\sim100\ \mathrm{Hz}$ , with merger and ringdown up to $\sim250\ \mathrm{Hz}$ .

If gravitational radiation is so weak, how was it ever detected?

By patience and leverage. The coupling is tiny — the strain at Earth is around $10^{-21}$ , a fractional length change smaller than a proton's width across a kilometre-scale interferometer. Two things make detection possible. First, matched filtering: because the waveform is known in advance, one coherently sums the signal across thousands of cycles, so the signal-to-noise ratio grows with accumulated phase, not instantaneous amplitude. Second, the source luminosity is not small — near merger it approaches $\sim10^{-2}\,c^5/G$ — so enough strain survives the $1/r$ dilution over hundreds of megaparsecs. Hulse–Taylor detected the effect indirectly, through the orbit's recoil, forty years before LIGO caught one directly.