Lagrange Point

In celestial mechanics and astrodynamics, Lagrange points (also designated Lagrangian points or libration points, labeled L₁ through L₅) are the five stationary equilibrium solutions of the circular restricted three-body problem (CR3BP). Within the orbital plane of two massive bodies (M₁ and M₂, such as the Sun and Earth, or the Sun and Jupiter) revolving in circular orbits about their common barycenter at constant angular velocity Ω, a test particle or spacecraft of negligible mass (m ≪ M₂ ≤ M₁) placed at rest at any of the five Lagrange points experiences a net gravitational attraction from M₁ and M₂ that provides the exact centripetal acceleration (Ω²r) required to orbit the barycenter with the identical orbital period as the two primary bodies.

The first three collinear equilibrium points (L₁, L₂, and L₃)—which lie along the straight line connecting M₁ and M₂—were discovered by the Swiss mathematician Leonhard Euler in 1767. Five years later, in his 1772 prize memoir Essai sur le problème des trois corps, the Italian-French mathematician Joseph-Louis Lagrange derived the complete five-point family, proving that two additional equilibrium points (L₄ and L₅) lie at the apexes of two equilateral triangles whose common base is the segment joining M₁ and M₂ (60° ahead of and 60° behind the secondary mass M₂ along its orbit). Today, natural populations of Trojan asteroids populate the stable L₄ and L₅ regions of Jupiter, Neptune, Mars, and Earth, while halo and Lissajous orbits around the Sun–Earth L₁ and L₂ points host flagship space observatories including SOHO, DSCOVR, Gaia, Euclid, and the James Webb Space Telescope (JWST).

🪐 Interactive Orbital Mechanics Simulation: Navigate three-body gravitational wells, Hill spheres, and multi-body transfer trajectories across the Solar System in Gravity Assist.

Equations of Motion in the Synodic Co-Rotating Reference Frame

Let two primary bodies of masses M₁ ≥ M₂ orbit their center of mass at constant separation R with Keplerian angular velocity:

Ω = √[ G ( M₁ + M₂ ) / R³ ]

Defining the dimensionless mass parameter μ^* = M₂ / (M₁ + M₂) ∈ (0, ½], the barycenter lies along the line of centers at distance x₁ = −*μ*^*R from M₁ and x₂ = (1 − μ^*)*R* from M₂. In a synodic reference frame (x, y, z) that rotates with angular velocity Ω = Ω k̂ so that M₁ and M₂ remain fixed on the x-axis, a test particle at distances r₁ and r₂ from M₁ and M₂ moves according to the rotating-frame equation of motion:

d²r / dt² + 2 Ω × ( *d*r / dt ) = − ∇*U*_eff(x, y, z)

where −2Ω × v_rot is the Coriolis acceleration and U_eff(x, y, z) is the effective (pseudogravitational) potential per unit mass combining the Newtonian gravitational potentials of M₁ and M₂ with the centrifugal potential:

U_eff(x, y, z) = − ( G M₁ / r₁ ) − ( G M₂ / r₂ ) − ½ Ω² ( x² + y² )

Taking the scalar dot product of the equation of motion with the rotating velocity *d*r/dt eliminates the perpendicular Coriolis force (v_rot · [Ω × v_rot] = 0) and yields a conserved quantity in the rotating frame known as the Jacobi integral (C_J) (which replaces the two-body vis-viva equation in the three-body problem):

C_J = − 2 U_eff(x, y, z) − v_rot² = Ω²(x² + y²) + ( 2 G M₁ / r₁ ) + ( 2 G M₂ / r₂ ) − v_rot² = constant

The five Lagrange points are the five critical points in the orbital plane (z = 0) where the gradient of the effective potential vanishes (∇*U*_eff = 0), so a particle placed there with zero rotating velocity (v_rot = 0) experiences zero net force in the rotating frame!

Physical Intuition and Algebraic Location of L₁, L₂, and L₃

Setting ∂U_eff/∂x = 0 along the x-axis (y = z = 0) yields a fifth-degree (quintic) polynomial with exactly one real root in each of the three intervals separated by M₁ and M₂:

1. The Inner Lagrange Point (L₁, between M₁ and M₂): Normally, an object orbiting inside Earth's orbit (r < 1 AU) has a shorter Keplerian period than 1 year. At L₁, however, Earth's gravity pulls outward (opposite to the Sun's pull), partially canceling solar gravity and slowing the orbital period down so that it equals exactly 1.000 year. For the small-mass ratio μ^* ≈ M₂ / M₁ ≪ 1, expanding the quintic equation to leading order shows that both L₁ and L₂ lie at the boundary of the secondary body's Hill sphere (r_H):

r_L1 ≈ r_L2 ≈ r_H = R · ( M₂ / [ 3 M₁ ] )^1/3 For the Sun–Earth system (R = 1.496 × 10⁸ km, M_⊕/M_☉ = 3.003 × 10⁻⁶), L₁ lies 1.491 million km (~0.010 AU, or ~3.9 lunar distances) sunward of Earth, providing an uninterrupted view of the Sun and the incoming solar wind (SOHO, ACE, WIND, DSCOVR, and India's Aditya-L1).

1. The Outer Lagrange Point (L₂, beyond M₂): Normally, an object orbiting outside Earth's orbit (r > 1 AU) has a longer Keplerian period than 1 year. At L₂, the gravitational pulls of the Sun and Earth add together in the same inward direction, increasing the centripetal force and speeding up the orbital angular velocity so the period matches 1.000 year. Located 1.501 million km (~0.010 AU) anti-sunward of Earth, Sun–Earth L₂ allows a single sunshield to block thermal radiation from the Sun, Earth, and Moon simultaneously—making it the premier site for cryogenic infrared and cosmic microwave background observatories (WMAP, Herschel, Planck, Gaia, Euclid, and JWST).
2. The Trans-Solar Lagrange Point (L₃, opposite M₂ behind M₁): Located on the far side of the primary body M₁ slightly outside the secondary's orbit at distance r_L3 ≈ R [1 − (7/12)(M₂/M₁)] from the barycenter.

Why L₄ and L₅ Are Stable Maxima of the Effective Potential (M₁ / M₂ > 24.96)

At L₄ and L₅, the test particle forms an equilateral triangle of side length r₁ = r₂ = R with both primary masses M₁ and M₂. Because the gravitational forces G M₁/R² and G M₂/R² are both inversely proportional to R² and directly proportional to M₁ and M₂, their vector resultant points directly through the barycenter of the system with magnitude G(M₁ + M₂)*r*_bary/R³ = Ω²r_bary, balancing the centrifugal force for any mass ratio!

Remarkably, plotting the effective potential surface U_eff(x, y) shows that L₁, L₂, and L₃ are saddle points (unstable along the radial x-direction), whereas L₄ and L₅ are local maxima (hilltops) of U_eff! In a non-rotating static system, a hilltop is always unstable. In the rotating frame, however, as a particle begins to slide down the potential hill away from L₄ or L₅ and picks up velocity v_rot, the Coriolis force (−2*m* Ω × v_rot) deflects the particle perpendicular to its motion, bending its trajectory into a closed kidney-bean-shaped (tadpole) libration orbit around L₄ or L₅ (exactly analogous to how the Coriolis force traps high-pressure anticyclonic winds on Earth or how a magnetic Penning trap confines charged particles).

Linearizing the equations of motion around L₄ and L₅ gives the characteristic equation λ⁴ + Ω²λ² + (27/4)*μ*^*(1 − μ^*)*Ω*⁴ = 0. For all four eigenvalues λ to be purely imaginary (non-growing oscillatory libration rather than exponential divergence exp(λt)), the discriminant must be positive:

1 − 27 μ^* ( 1 − μ^* ) > 0 ⇒ μ^* < ( 1 − √[23/27] ) / 2 ≈ 0.03852 ⇒ M₁ / M₂ > ( 25 + √621 ) / 2 ≈ 24.9599

Known as Gascheau's (or Routh's) critical mass ratio (derived by G. Gascheau in 1843), this condition M₁/M₂ > 24.96 is satisfied by the Sun–Jupiter (M_☉/M_J ≈ 1,047), Sun–Earth (~333,000), and Earth–Moon (M_⊕/M_☾ ≈ 81.3) systems! Consequently, Jupiter hosts over 12,000 cataloged Trojan asteroids (visited by NASA's Lucy mission), and Saturn's moons Tethys and Dione host small co-orbital Trojan moons (Telesto, Calypso, Helene, and Polydeuces) in their L₄ and L₅ points.

If Sun–Earth L₁ and L₂ are unstable saddle points, how do spacecraft like JWST stay there?

The three collinear Lagrange points (L₁, L₂, L₃) possess one positive real eigenvalue λ₊ > 0 (an unstable exponential divergence mode, x(t) ∝ exp(λ₊t), with an e-folding time of τ = 1/λ₊ ≈ 23 days at Sun–Earth L₁ and L₂) alongside two pairs of purely imaginary oscillatory eigenvalues (one in-plane and one out-of-plane). Spacecraft do not sit statically at the geometric point L₂; instead, they are injected onto the three-dimensional periodic or quasi-periodic Halo orbit or Lissajous orbit family (discovered by Robert Farquhar in the 1960s) with an amplitude of ~250,000 to 800,000 km around L₂. This keeps the James Webb Space Telescope (JWST) out of Earth's shadow so its solar arrays stay constantly illuminated, and requires only tiny station-keeping thruster burns of Δ*v* ≈ 2 to 4 m s⁻¹ per year every few weeks to damp out the unstable eigenvalue mode!

Why did China place the Queqiao relay satellite in a Halo orbit around the Earth–Moon L₂ point?

Because the Moon is locked in a 1:1 synchronous orbital resonance (spin–orbit lock) with Earth, the lunar far side (including the South Pole–Aitken Basin) never faces Earth, blocking direct radio line-of-sight communication with lunar far-side landers and rovers (Chang'e 4 and Chang'e 6). By placing the Queqiao relay satellite into a ~13,000 km × 42,000 km Halo orbit around the Earth–Moon L₂ point (~64,500 km behind the Moon), the satellite circles wide enough around the lunar disk to maintain simultaneous, uninterrupted line-of-sight radio links with both the lunar far-side surface and ground tracking stations on Earth!

What is the "Interplanetary Transport Network" connecting Lagrange points across the Solar System?

Each unstable collinear Lagrange point (L₁ and L₂) possesses four-dimensional stable and unstable invariant manifold tubes in phase space—trajectories that naturally wind onto or unwind off of Halo orbits with zero propellant cost. When the invariant manifold tubes extending from Sun–Earth L₁/L₂ intersect those of Earth–Moon L₁/L₂ or those of neighboring planets, a spacecraft can transfer along these low-energy heteroclinic corridors (used by Genesis, ARTEMIS, Hiten, and Danuri) for a fraction of the Δ*v* required by a classic Hohmann transfer orbit, at the cost of longer flight times.

1. Lagrange, J.-L. (1772). "Essai sur le problème des trois corps." Prix de l'Académie Royale des Sciences de Paris, 9, 229–331 (In Œuvres de Lagrange, Vol. 6). https://gallica.bnf.fr/ark:/12148/bpt6k229225d/f231.item
2. Szebehely, V. (1967). Theory of Orbits: The Restricted Problem of Three Bodies. Academic Press. https://doi.org/10.1016/B978-0-12-395732-0.X5001-6
3. Murray, C. D., & Dermott, S. F. (1999). Solar System Dynamics. Cambridge University Press. https://doi.org/10.1017/CBO9781139174817
4. Koon, W. S., Lo, M. W., Marsden, J. E., & Ross, S. D. (2011). Dynamical Systems, the Three-Body Problem and Space Mission Design. Marsden Books. ISBN 978-0615240954