Schwarzschild and Kerr Black Holes

A black hole is the one place in physics where geometry becomes destiny. General relativity permits vacuum solutions — spacetimes solving Einstein's equations with zero stress–energy — in which curvature survives even where no matter remains. The first and simplest, written by Karl Schwarzschild in 1916 within weeks of the field equations and while he lay ill on the Eastern Front, describes the entire exterior of any static, spherical mass with a single number: the mass M. Birkhoff's theorem makes it unique — every spherically symmetric vacuum region is a slice of Schwarzschild spacetime and is automatically static — so a pulsating or collapsing spherical star has exactly the external field of a point mass, and gravitational monopole radiation cannot exist. Compress that mass inside its Schwarzschild radius $r_s = 2GM/c^2$ and the surface $r=r_s$ becomes a one-way membrane: a null hypersurface through which light and matter fall in but never climb out. The metric misbehaves there, but that is a fault of the coordinates — Eddington–Finkelstein and Kruskal–Szekeres charts cross smoothly and curvature stays finite. The true singularity hides at $r=0$ , where the Kretschmann scalar diverges. This is the generic endpoint of gravitational collapse once degeneracy and nuclear pressure fail — the sequel to white dwarfs, the Chandrasekhar limit, neutron stars, and the Tolman–Oppenheimer–Volkoff equation. Real stars rotate, and angular momentum survives collapse, so real black holes spin. The Kerr solution (Roy Kerr, 1963) is the unique stationary, axisymmetric, asymptotically flat vacuum with a regular horizon, carrying two numbers — mass M and spin $a=J/Mc$ . Its structure is richer: the horizon and the surface of infinite redshift split apart, and between them lies the ergosphere, a shell where frame dragging lets nothing stand still and from which rotational energy can be mined. With charge these metrics realize the no-hair theorem, and their signatures — accretion spectra, shadows, and merger ringdowns — now test the vacuum of general relativity to percent precision.

Key takeaways

The Schwarzschild metric is the unique spherically symmetric vacuum solution of Einstein's equations — a non-rotating, uncharged black hole fixed by mass M alone, with an event horizon at $r_s = 2GM/c^2 \approx 2.95$ km per solar mass.
The Kerr metric adds rotation: a spinning hole set by M and spin parameter $a=J/Mc$ , with an outer horizon $r_+$ and, outside it, an ergosphere where no observer can remain static.
The metric blows up at the horizon only in bad coordinates — curvature invariants stay finite there; the genuine curvature singularity is at $r=0$ (a point for Schwarzschild, a ring for Kerr).
Light circles a Schwarzschild hole at the photon sphere $3GM/c^2$ , and the innermost stable circular orbit sits at $6GM/c^2$ , fixing the inner edge of accretion disks and a maximum radiative efficiency of about 5.7 percent.
The Kerr ergosphere permits energy extraction — the Penrose process and its field version, the Blandford–Znajek mechanism — draining spin to power relativistic jets, with the spin bounded by $a \le GM/c^2$ .
By the no-hair theorem an isolated equilibrium black hole is fully specified by M, J, and charge Q; astrophysical charge neutralizes, so real holes are two-parameter Kerr.
Key idea: the exterior field of an isolated compact object is fixed by at most three numbers — mass, spin, and charge — through the Schwarzschild and Kerr–Newman vacuum solutions; horizons are causal boundaries that coordinates can hide, while true singularities are curvature divergences locked inside.

The core insight

The exterior gravitational field of an isolated compact object is fixed by at most three numbers — mass, spin, and charge — through the Schwarzschild and Kerr–Newman vacuum solutions; horizons are coordinate-removable causal boundaries, while true singularities are curvature divergences hidden inside.

Where this lives today

Black-hole metrics have moved from blackboard to instrument. The Event Horizon Telescope resolved the photon-ring shadows of M87 (2019) and Sgr A (2022), matching the $\sqrt{27}\,GM/c^2$ prediction. LIGO–Virgo–KAGRA now catches black-hole mergers routinely; the post-merger ringdown, a superposition of Kerr quasinormal modes whose frequencies depend only on final mass and spin, directly tests the no-hair theorem, and detecting a second mode ("black-hole spectroscopy") is a headline goal for the space-based LISA. X-ray astronomers weigh spins from the ISCO-truncated disk continuum and the relativistically broadened iron Kα line, whose red wing is sculpted by redshift near $r_+$ ; measured spins cluster near the $a\approx0.998$ accretion ceiling. GRAVITY has tracked the pericenter precession of the star S2 and hot spots orbiting near the ISCO of Sgr A*. And the Blandford–Znajek mechanism — the electromagnetic cousin of the Penrose process, with jet power $\propto\Omega_H^2$ — remains the leading engine for the relativistic jets of active galactic nuclei and gamma-ray bursts. Open problems persist: the information paradox, the fate of the inner Cauchy horizon, and whether cosmic censorship holds in general.

The physical picture

Schwarzschild spacetime is the frozen exterior of a spherical mass: distant clocks run fast, deep clocks run slow, radial rulers stretch. Light climbing out loses energy, so a source lowered toward $r_s$ reddens and dims without limit — from afar the object appears to freeze at the horizon, the origin of the old name "frozen star." Yet nothing local marks the crossing: an infalling observer sails through $r_s$ in finite proper time, feeling only tides, and meets ruin only near $r=0$ . The geometry's special radii are a nested set — the horizon, the photon sphere where light can orbit, and the ISCO inside which matter must plunge. The first figure lays them out to scale for a one-solar-mass hole, so the ratios $1 : 1.5 : 3$ read off directly.

Figure

Anatomy of a Schwarzschild black hole to scale for one solar mass: the event horizon at r_s = 2GM/c² = 2.95 km, the photon sphere at 3GM/c² = 4.43 km where light can orbit, and the innermost stable circular orbit at 6GM/c² = 8.86 km, with the curvature singularity at r = 0

Rotation changes the picture qualitatively. A spinning mass drags spacetime around with it — frame dragging — so near a Kerr hole the local inertial frames are forced to co-rotate. Cross the static limit and the dragging becomes irresistible: to hold still against the distant stars you would have to move faster than light, so no static observer exists. This shell between the static limit and the horizon is the ergosphere, and because the Killing vector $\partial_t$ turns spacelike inside it, a particle there can carry negative energy-at-infinity. Drop that fragment down the hole and the escaping piece leaves with more energy than went in — you have mined the hole's spin. The second figure is a poloidal cut through a rapidly spinning Kerr hole ( $a = 0.9\,GM/c^2$ ), with the oblate ergosphere bulging to $2GM/c^2$ at the equator while pinching onto the horizon at the poles.

Figure

Poloidal cross-section of a Kerr black hole with spin a = 0.9 GM/c²: the outer event horizon is at r₊ = 1.44 GM/c², the static limit bulges to 2 GM/c² at the equator and touches the horizon at the poles, and the ergosphere is the wedge between them where no observer can stay at rest

Both horizons are null surfaces — woven from light rays that neither escape nor fall in but hover forever — which is why they act as one-way membranes, and Hawking's area theorem guarantees their total area never decreases, the geometric seed of black-hole thermodynamics. The two orbital radii of the first figure follow from a single fact: the relativistic effective potential carries an extra attractive $1/r^3$ term absent from Newton's, which forbids any stable circle inside $6GM/c^2$ .

The core equations

The Schwarzschild metric is the master object: the complete exterior geometry of any non-rotating spherical mass, in coordinates adapted to a static distant observer. Everything about a non-spinning black hole reads off from these four coefficients.

\mathrm{d}s^2 = -\left(1-\frac{2GM}{c^2 r}\right)c^2\,\mathrm{d}t^2 + \left(1-\frac{2GM}{c^2 r}\right)^{-1}\mathrm{d}r^2 + r^2\left(\mathrm{d}\theta^2 + \sin^2\theta\,\mathrm{d}\phi^2\right) \qquad (1)

where: $M$ — total (ADM) mass; $r$ — areal radial coordinate, so a sphere of coordinate $r$ has proper area $4\pi r^2$ ; $t$ — coordinate time of a static observer at infinity; $\theta,\phi$ — polar and azimuthal angles; $G,c$ — Newton's constant and the speed of light. Reading: $g_{tt}=-(1-2GM/c^2r)$ vanishes at $r_s$ (infinite redshift) and $g_{rr}$ diverges there — the coordinate horizon. Setting that vanishing factor to zero defines the length that scales with mass — the Schwarzschild radius, the horizon of a non-rotating hole and the yardstick for every black-hole size.

r_s = \frac{2GM}{c^2} \approx 2.95\ \mathrm{km}\left(\frac{M}{M_\odot}\right) \qquad (2)

where: $r_s$ — event-horizon radius of a non-rotating black hole; $M_\odot$ — solar mass, $1.989\times10^{30}$ kg. Reading: the horizon grows linearly with mass — a stellar hole is city-sized, a billion-solar-mass hole spans the inner Solar System. To prove the horizon is a coordinate artifact, evaluate a curvature scalar no coordinate change can fool — the Kretschmann invariant, finite at $r_s$ but divergent at $r=0$ .

K \equiv R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma} = \frac{48\,G^2 M^2}{c^4 r^6} \qquad (3)

where: $R_{\mu\nu\rho\sigma}$ — Riemann curvature tensor; $K$ — quadratic curvature invariant. Reading: at $r=r_s$ this equals $3c^8/(4G^4M^4)$ , large but finite; as $r\to0$ it diverges — the one true singularity. Spin the source and the geometry becomes the Kerr metric in Boyer–Lindquist coordinates — stationary and axisymmetric, with an off-diagonal $\mathrm{d}t\,\mathrm{d}\phi$ term encoding frame dragging.

\mathrm{d}s^2 = -\left(1-\frac{r_s r}{\Sigma}\right)c^2\mathrm{d}t^2 - \frac{2 r_s r\, a c\sin^2\theta}{\Sigma}\mathrm{d}t\,\mathrm{d}\phi + \frac{\Sigma}{\Delta}\mathrm{d}r^2 + \Sigma\,\mathrm{d}\theta^2 + \left(r^2+a^2+\frac{r_s r a^2\sin^2\theta}{\Sigma}\right)\sin^2\theta\,\mathrm{d}\phi^2 \qquad (4)

where: $a=J/(Mc)$ — spin parameter, dimensions of length; $\Sigma = r^2 + a^2\cos^2\theta$ ; $\Delta = r^2 - r_s r + a^2$ — horizon function; $r_s=2GM/c^2$ ; $J$ — total angular momentum. Reading: the cross term $g_{t\phi}$ drags frames; it vanishes as $a\to0$ and (4) collapses to (1). The horizons of a Kerr hole are the zeros of $\Delta$ , and requiring real zeros caps the spin.

r_\pm = \frac{GM}{c^2} \pm \sqrt{\left(\frac{GM}{c^2}\right)^2 - a^2} \qquad (5)

where: $r_+$ — outer (event) horizon; $r_-$ — inner (Cauchy) horizon. Reading: they merge at $a=GM/c^2$ (extremal Kerr); beyond that the root is imaginary and no horizon exists. The static limit — the surface of infinite redshift where $g_{tt}=0$ — lies outside the horizon except at the poles; the shell between them is the ergosphere.

r_{\rm ergo}(\theta) = \frac{GM}{c^2} + \sqrt{\left(\frac{GM}{c^2}\right)^2 - a^2\cos^2\theta} \qquad (6)

where: $r_{\rm ergo}$ — outer boundary of the ergosphere; $\theta$ — polar angle from the spin axis. Reading: at the poles it equals $r_+$ ; at the equator it reaches its maximum bulge $2GM/c^2$ . Orbits around a Schwarzschild hole obey an effective potential carrying one term Newton never wrote — the relativistic $-GML^2/(c^2r^3)$ that destabilizes tight orbits.

V_{\rm eff}(r) = -\frac{GM}{r} + \frac{L^2}{2r^2} - \frac{GM L^2}{c^2 r^3} \qquad (7)

where: $V_{\rm eff}$ — effective potential per unit mass for a massive test particle; $L$ — conserved specific angular momentum; the $1/r^3$ term is the purely relativistic correction. Reading: negative and steep at small $r$ , it erases the centrifugal barrier and closes off stable close orbits. The extrema of (7) fix the photon sphere for light and the ISCO for matter, while (9) gives the two Kerr horizon numbers that enter thermodynamics and energy extraction.

r_{\rm ph} = \frac{3GM}{c^2}, \qquad r_{\rm ISCO} = \frac{6GM}{c^2} \qquad (8)

\Omega_H = \frac{a c}{r_+^2 + a^2}, \qquad A = 4\pi\left(r_+^2 + a^2\right) \qquad (9)

where: $r_{\rm ph}$ — unstable circular photon orbit; $r_{\rm ISCO}$ — innermost stable circular orbit for matter; $\Omega_H$ — rigid angular velocity of the Kerr horizon; $A$ — proper area of the outer horizon. Reading: the ISCO sets the disk's inner edge and efficiency, while $\Omega_H$ and $A$ enter the Penrose bound and area theorem.

Derivation AThe photon sphere and ISCO from the geodesic effective potential

The goal here is to derive the two orbital landmarks of a Schwarzschild black hole — the photon sphere at $3GM/c^2$ and the innermost stable circular orbit at $6GM/c^2$ — directly from the geodesic equation, and to show both follow from the shape of one effective potential whose relativistic $1/r^3$ term has no Newtonian analogue. We use only the two Killing symmetries and the four-velocity normalization, then read circular orbits off the potential's first derivative and their stability off the second.

Step 1. For equatorial motion ( $\theta=\pi/2$ ) the Killing vectors $\partial_t$ and $\partial_\phi$ give two conserved quantities, the specific energy $\tilde E = E/mc^2$ and angular momentum $L$ ; the four-velocity normalization $u^\mu u_\mu = -c^2$ then yields a radial equation $\left(\mathrm{d}r/\mathrm{d}\tau\right)^2 = \tilde E^2 c^2 - \left(1-r_s/r\right)\left(c^2 + L^2/r^2\right)$ . Why: conservation laws collapse the geodesic equations into a single one-dimensional energy problem, as in Newtonian orbits — but with a relativistically modified potential.

Step 2. Circular orbits sit at extrema of the effective potential, $\mathrm{d}V_{\rm eff}/\mathrm{d}r=0$ . Carrying this out for a massive particle gives $L^2 = \dfrac{GM r^2}{r - 3GM/c^2}$ , which is positive and finite only for $r>3GM/c^2$ . Why: a real, finite angular momentum is required for a genuine circular orbit, so this already tells us no massive circular orbit of any kind exists inside $3GM/c^2$ .

Step 3. Take the photon limit by dropping the rest-mass term ( $c^2\to0$ ); the required $L$ diverges as $r\to3GM/c^2$ , leaving a single unstable circular null orbit — the photon sphere. Why: light has no rest mass to anchor a family of orbits, so only the marginal radius survives, and its instability is why the photon ring is a thin bright edge, not a filled disk.

Step 4. Stability needs the second derivative: the innermost stable orbit is where the potential minimum flattens into an inflection, $\mathrm{d}^2 V_{\rm eff}/\mathrm{d}r^2=0$ . Solving this with the circular-orbit condition gives $r_{\rm ISCO}=6GM/c^2$ , where $\tilde E=\sqrt{8/9}$ and the binding energy is $1-\sqrt{8/9}\approx0.0572$ . Why: inside the ISCO the minimum has vanished, so any inward nudge plunges — and 5.7 percent is the maximum rest-mass fraction a thin disk can radiate around a non-rotating hole, rising to about 42 percent for extremal Kerr.

Go deeper · Derivation B · 1 minKerr horizons and the extremal spin bound

Derivation BKerr horizons and the extremal spin bound

The aim is to locate the horizons of a rotating black hole and to prove that spin cannot exceed a sharp bound, $a\le GM/c^2$ , without destroying the horizon altogether. The horizons are the surfaces where $g_{rr}=\Sigma/\Delta$ diverges, i.e. where $\Delta$ vanishes; since all curvature invariants stay finite there, they are true null horizons, not curvature singularities. We solve the quadratic and read the reality condition off its discriminant.

Step 1. Set the horizon function to zero: $\Delta = r^2 - \dfrac{2GM}{c^2}r + a^2 = 0$ , a plain quadratic in $r$ whose roots are the candidate horizons. Why: $\Delta$ multiplies the inverse radial metric, so its zeros are exactly where an ingoing light cone tips fully inward — the defining property of a horizon.

Step 2. The quadratic formula gives $r_\pm = \dfrac{GM}{c^2}\pm\sqrt{\left(GM/c^2\right)^2 - a^2}$ , the outer event horizon $r_+$ and inner Cauchy horizon $r_-$ . Why: two nested null surfaces, and the strange causal region between them is why the Kerr interior differs so from Schwarzschild's.

Step 3. Real roots demand a non-negative discriminant, $(GM/c^2)^2 - a^2 \ge 0$ , i.e. $a\le GM/c^2$ , equivalently $J\le GM^2/c$ . Why: the discriminant condition is the entire content of the extremal spin limit — spin any faster and the square root goes imaginary.

Step 4. At equality the two horizons merge at $r_+=r_-=GM/c^2$ (extremal Kerr, zero surface gravity and zero Hawking temperature); beyond it $\Delta$ has no real root and the ring singularity would be naked. Why: a naked singularity would let the unpredictable interior reach the outside universe, which cosmic censorship forbids — and astrophysical spin-up saturates near $a\approx0.998\,GM/c^2$ , short of extremal.

Go deeper · Derivation C · 1 minThe Penrose process and the irreducible mass

Derivation CThe Penrose process and the irreducible mass

The purpose here is to show precisely how energy is extracted from a spinning black hole, and why the extraction is nonetheless bounded. The key is that inside the ergosphere the Killing vector $\xi=\partial_t$ becomes spacelike, so the conserved energy $E=-p_\mu\xi^\mu$ of a physical, future-directed particle can be negative. We split a particle there, feed the negative-energy fragment to the hole, and track the response down to the irreducible-mass floor set by the area theorem.

Step 1. Send a particle of energy $E_0$ into the ergosphere and let it break in two, $E_0 = E_1 + E_2$ , arranged so that fragment 1 has $E_1<0$ and falls through the horizon while fragment 2 escapes. Why: only inside the ergosphere is a negative conserved energy kinematically allowed, so this is the one region where the trick works at all.

Step 2. Energy conservation gives $E_2 = E_0 - E_1 > E_0$ — the escaping fragment carries away more than the whole particle brought in. Why: the surplus is not created from nothing; it is drawn from the hole, which loses exactly $E_1$ of mass-energy.

Step 3. Absorbing the negative-energy fragment lowers both the hole's mass and its spin, $\delta M<0$ and $\delta J<0$ , linked by $\delta M = (\Omega_H/c^2)\,\delta J$ ; the energy is paid out of rotation. Why: $\Omega_H$ enforces that you can only extract energy by also removing spin, which is why a non-rotating hole yields nothing.

Step 4. The horizon area can never decrease (Hawking), and area fixes the irreducible mass through $M^2 = M_{\rm irr}^2 + \dfrac{J^2 c^2}{4G^2 M_{\rm irr}^2}$ , so extraction halts when the spin is gone; extremal Kerr yields $1-1/\sqrt2\approx29\%$ . Why: $M_{\rm irr}$ is the true floor — it is $\sqrt{A}$ dressed as a mass, and since $A$ cannot shrink, no process pushes $M$ below it.

Limiting cases

The Kerr family contains Schwarzschild and flat Minkowski as limits, and its horizon structure hinges entirely on the spin:

Case	Result	Meaning
No spin, $a\to0$	$\mathrm{d}s^2 \to -\left(1-\tfrac{2GM}{c^2 r}\right)c^2\mathrm{d}t^2 + \left(1-\tfrac{2GM}{c^2 r}\right)^{-1}\mathrm{d}r^2 + r^2\mathrm{d}\Omega^2$	Kerr collapses to Schwarzschild; the ergosphere shrinks onto the horizon.
No mass, $M\to0$	$\mathrm{d}s^2 \to -c^2\mathrm{d}t^2 + \mathrm{d}r^2 + r^2\mathrm{d}\Omega^2$	Flat Minkowski spacetime; curvature vanishes everywhere.
Extremal Kerr, $a=GM/c^2$	$r_+ = r_- = \dfrac{GM}{c^2}$	Inner and outer horizons merge; maximal spin. Higher spin gives a naked ring singularity, forbidden by cosmic censorship.
Far field, $r\gg r_s$	$g_{tt} \approx -\left(1 + \dfrac{2\Phi_N}{c^2}\right),\ \Phi_N=-\dfrac{GM}{r}$	Recovers Newtonian gravity, plus a gravitomagnetic Lense–Thirring term $\propto J$ for Kerr.

Worked examples

1 · Sizing three black holes (confidence builder)

Use $r_s=2.95\,\mathrm{km}\,(M/M_\odot)$ and the ratios of equation (8) to size a 10 M⊙ stellar-mass hole and the M87 supermassive hole ( $6.5\times10^9\,M_\odot$ ).*

Step 1. Stellar hole, $M=10\,M_\odot$ : horizon $r_s = 2.95\times10 = \mathbf{29.5\ km}$ , photon sphere $1.5\,r_s = 44.3$ km, ISCO $3\,r_s = \mathbf{88.6\ km}$ . A black hole heavier than the Sun would sit inside a mid-sized city.

Step 2. Supermassive hole, $M=6.5\times10^9\,M_\odot$ : $r_s = 2.95\ \mathrm{km}\times6.5\times10^9 = 1.92\times10^{10}$ km. With $1\ \mathrm{AU}=1.496\times10^8$ km this is $r_s = \mathbf{128\ AU}$ , and the ISCO at $3\,r_s = \mathbf{385\ AU}$ — nine times the size of Neptune's orbit.

Step 3. Every length scales linearly with M, so the two holes differ in size by the same factor $6.5\times10^8$ as in mass. A 10 M⊙ hole is city-sized (r_s = 29.5 km); M87* spans the inner Solar System (r_s = 128 AU).

2 · The shadow of Sagittarius A* (fully worked)

A Schwarzschild hole casts a dark shadow of angular diameter $2\sqrt{27}\,GM/(c^2 D)$ , set by the critical photon impact parameter $b_c=\sqrt{27}\,GM/c^2$ . Predict the shadow of Sgr A ( $M=4.15\times10^6\,M_\odot$ , $D=8.15$ kpc) and compare with the EHT.*

Step 1. Gravitational radius: $r_g = GM/c^2 = 1.48\ \mathrm{km}\,(M/M_\odot)\times4.15\times10^6 = \mathbf{6.13\times10^6\ km}$ , about $0.041$ AU — comparable to Mercury's orbit.

Step 2. Shadow linear diameter: $2\sqrt{27}\,r_g = 2\times5.196\times6.13\times10^6 = 6.37\times10^7$ km. Distance: $D=8.15\ \mathrm{kpc}\times3.086\times10^{16}\ \mathrm{km/kpc} = 2.52\times10^{17}$ km.

Step 3. Angular diameter: $\theta = 6.37\times10^7 / 2.52\times10^{17} = 2.53\times10^{-10}$ rad $= 2.53\times10^{-10}\times2.063\times10^{11}\,\mu\mathrm{as/rad} = \mathbf{52\ \mu as}$ .

Step 4. The 2022 EHT image measured a ring diameter of $51.8\pm2.3\ \mu$ as — striking agreement, since spin and inclination shift the value by only a few percent. Predicted shadow ≈ 52 μas, matching the EHT and confirming a strong-field Kerr geometry.

3 · Why quasars outshine galaxies

Accretion releases the ISCO binding energy: $L = \eta\,\dot M c^2$ . Take the Schwarzschild efficiency $\eta = 1-\sqrt{8/9}=0.057$ and one solar mass per year, and compare with fusion ( $\eta_{\rm fus}\approx0.007$ ).

Step 1. Rate: $\dot M = M_\odot/\mathrm{yr} = 1.989\times10^{30}/3.156\times10^{7} = 6.30\times10^{22}$ kg/s.

Step 2. Luminosity: $L = 0.057\times6.30\times10^{22}\times(2.998\times10^8)^2 = \mathbf{3.2\times10^{38}\ W}=3.2\times10^{45}$ erg/s — squarely in the quasar range.

Step 3. Accretion at 5.7 percent is about eight times more efficient than fusion's 0.7 percent, and extremal-Kerr accretion at 42 percent is sixty times more efficient. Black-hole accretion, not fusion, is the most efficient sustained energy source in the universe — the engine behind quasars and AGN.

Physics of the portrait

The black hole is the rare case where an exact solution came before anyone believed the object was real — and the two men who wrote the defining metrics did so nearly half a century apart.

Origins · 1 minKarl Schwarzschild and how this was discovered

Karl Schwarzschild

1873–1916 · the exterior vacuum solution found on the Eastern Front, 1916

Karl Schwarzschild was already a distinguished astronomer — director of the Potsdam Observatory, a pioneer of stellar photometry — when the First World War sent him, in his forties, to the Russian front to compute artillery trajectories. There, within weeks of Einstein presenting the field equations in November 1915, he found their first exact non-trivial solution: the complete vacuum geometry outside a spherical mass, in a letter Einstein read to the Prussian Academy in January 1916. Einstein was astonished a closed-form solution existed at all. Schwarzschild contracted pemphigus at the front and died in May 1916, months after his geometry was published, never knowing it described what we now call a black hole. For decades the surface $r=2GM/c^2$ was dismissed as a pathology — the "Schwarzschild singularity" — until Eddington, Finkelstein, and Kruskal showed in the 1950s it was a smooth horizon in disguise. The rotating generalization waited until 1963, when the New Zealand mathematician Roy Kerr, wrestling with algebraically special solutions, extracted the stationary axisymmetric metric that bears his name — the one that describes the spinning holes nature makes. Subrahmanyan Chandrasekhar, who computed the mass limit that forces the collapse, called it the most important exact result in the history of general relativity, and marvelled that "black holes are the most perfect macroscopic objects there are in the universe" — built from spacetime alone, fixed by mass and spin.

History

1915

— Einstein presents the field equations of general relativity

1916

— Schwarzschild solves them for a point mass, giving the first black-hole metric

1939

— Oppenheimer and Snyder model the continued gravitational collapse of a star

1963

— Kerr finds the rotating, axisymmetric vacuum solution

1969

— Penrose proposes ergosphere energy extraction and the cosmic censorship conjecture

2019

— the Event Horizon Telescope images the shadow of M87*

Going further

Texts. Carroll and Ostlie, An Introduction to Modern Astrophysics, 2nd ed., Ch. 17 and 24 · Misner, Thorne and Wheeler, Gravitation, Ch. 25, 31, 33 · Wald, General Relativity, Ch. 6 and 12 · Chandrasekhar, The Mathematical Theory of Black Holes · Shapiro and Teukolsky, Black Holes, White Dwarfs, and Neutron Stars, Ch. 12. Key numerical anchors to keep in mind: $r_s\approx2.95$ km and $r_g=1.48$ km per solar mass; the Kerr horizons are the roots of $\Delta=0$ with the spin bound $a\le GM/c^2$ ; the effective-potential extrema give $r_{\rm ph}=3GM/c^2$ and $r_{\rm ISCO}=6GM/c^2$ with binding energy $1-\sqrt{8/9}\approx0.0572$ ; the Kretschmann scalar at the horizon equals the finite value $3c^8/(4G^4M^4)$ ; and the Sgr A* shadow diameter $2\sqrt{27}\,GM/(c^2D)\approx52\,\mu$ as agrees with the EHT measurement of $51.8\pm2.3\,\mu$ as.

Prerequisites. Tolman–Oppenheimer–Volkoff Equation · Neutron Stars · Chandrasekhar Mass Limit · Equation of State of Dense Matter. Leads to. Magnetars · Accretion Disks and Efficiency · Gravitational Waves from Mergers.

Ask the tutor: Want to walk through why the ISCO sits exactly at $6GM/c^2$ , or how the Sgr A* shadow pins down the black hole's mass — just ask.

Why is r = r_s not a real singularity if the metric blows up there?

Because "the metric blows up" is a statement about coordinates, not geometry. The component $g_{rr}=(1-r_s/r)^{-1}$ diverges at $r_s$ , but that is like longitude lines crowding to a useless point at the Earth's pole — the sphere is smooth there, the map just fails. The test that cannot be fooled by coordinates is a curvature scalar: the Kretschmann invariant $K=48G^2M^2/(c^4r^6)$ is finite at $r=r_s$ and diverges only at $r=0$ . Kruskal–Szekeres or Eddington–Finkelstein coordinates remove the apparent problem entirely, and an infalling observer crosses $r_s$ in finite proper time feeling nothing dramatic. The horizon is a causal boundary — a point of no return — not a place where physics ends.

What is the difference between the ergosphere and the event horizon?

The event horizon at $r_+$ is the true one-way membrane: cross it and you cannot come back, because it is a null surface where $\Delta=0$ and all future-directed paths lead inward. The ergosphere lies entirely outside the horizon, between $r_+$ and the static limit where $g_{tt}=0$ . Inside it, frame dragging is so strong that you cannot remain at rest relative to distant stars — you are forced to co-rotate — but you can still escape to infinity, so it is not a trap. Its importance is energetic: because the time-translation Killing vector turns spacelike there, particles can carry negative energy-at-infinity, which is what makes Penrose and Blandford–Znajek energy extraction possible.

How much energy can you actually extract from a spinning black hole?

Up to about 29 percent of the total mass-energy for a maximally spinning (extremal) Kerr hole, and less for slower spins. The bound comes from the irreducible mass $M_{\rm irr}$ , defined through the horizon area by $M^2 = M_{\rm irr}^2 + J^2c^2/(4G^2M_{\rm irr}^2)$ . Since Hawking's area theorem forbids the area — and therefore $M_{\rm irr}$ — from ever decreasing, you can only mine the difference between the full mass and $M_{\rm irr}$ , which is pure rotational energy; for extremal Kerr that is $1-1/\sqrt2\approx0.293$ . Once the spin is gone the hole is Schwarzschild and yields nothing more. The astrophysically important version is not the mechanical Penrose process but its electromagnetic analogue, the Blandford–Znajek mechanism, which taps the same reservoir through magnetic fields threading the horizon.

Why can't a black hole spin faster than a = GM/c²?

Because faster spin would erase the horizon. The horizons are the roots of $\Delta = r^2 - (2GM/c^2)r + a^2=0$ , and real roots require the discriminant $(GM/c^2)^2 - a^2\ge0$ , i.e. $a\le GM/c^2$ . At $a=GM/c^2$ (extremal Kerr) the horizons merge; beyond it the discriminant goes negative, no horizon exists, and the ring singularity would be exposed — a "naked" singularity that Penrose's cosmic censorship conjecture forbids. Reassuringly the physics cooperates: as a hole spins up by accretion, radiation carrying away angular momentum grows more efficient, and the spin saturates at $a\approx0.998\,GM/c^2$ — the Thorne limit — before extremality is reached.

Do real astrophysical black holes really have no hair?

To excellent approximation, yes. The no-hair theorem states that a stationary, isolated black hole in equilibrium is fully characterized by three numbers — mass M, angular momentum J, and charge Q — the Kerr–Newman family; every higher multipole is a fixed function of these. Any lumps or memory of the progenitor star are radiated away in gravitational waves as the hole settles. Charge is essentially always negligible: a charged hole in a plasma attracts opposite charges and neutralizes within moments, so $Q\approx0$ and real holes are two-parameter Kerr. The one caveat is "isolated and stationary" — a hole surrounded by a massive disk, or freshly formed and still ringing, carries transient structure until it relaxes.

Is the Kerr singularity really a ring, and could you pass through it?

In Boyer–Lindquist coordinates the Kerr curvature singularity sits at $\Sigma=0$ , which for $a\ne0$ is not a point but a ring of coordinate radius $a$ in the equatorial plane. Formally the vacuum solution can be extended through the ring into a region of negative $r$ containing closed timelike curves — features that have fascinated mathematicians and science-fiction writers alike. But these are almost certainly not physical: the inner Cauchy horizon at $r_-$ is violently unstable, and infalling radiation is expected to build into a curvature singularity there (mass inflation), sealing off the exotic interior. The clean, traversable ring is an artifact of the idealized eternal-vacuum solution, not a real collapsed star.