Gravitation & Satellite Orbits · Guide · glidepathscience.org
GuideGravitation & Satellite Orbits
Contents
This Guide explains the ideas and equations behind every Walkthrough step, Concept Challenge question and Predict round, in the order of a textbook chapter on gravitation. Part 1 prepares the Intro track and Part 2 the Advanced track; Part 1 takes about eighteen minutes to read and Part 2 about fifteen. Each section ends with a worked example that is not one of the sim’s own problems.
Part 1 · The essentials
Falling around the world
Throw a ball sideways and it curves down to the ground. Throw it harder and it lands farther away. Isaac Newton asked what happens if you keep going. In A Treatise of the System of the World (published in 1728, after his death) he imagined a cannon on a mountain so tall that its top stands above the air, firing horizontally with more and more powder. Each ball falls, but the faster ones travel so far before they land that the Earth’s surface has curved away beneath them. Fast enough, the ball falls all the way around and comes back to the mountain from behind. It is still falling the whole time; it simply keeps missing the ground.
A satellite, the Moon included, is a projectile fast enough that the ground curves away as fast as it falls. To say how fast, we need gravity’s strength at any distance from the Earth.
Newton’s law of universal gravitation
Every mass attracts every other mass. Newton proposed that the pull between two point masses, or between two spheres, is proportional to each mass and falls off as the square of the distance between their centers:
$F = G\dfrac{Mm}{r^2}$
Here M and m are the two masses, r is the distance from center to center, and G is the gravitational constant, 6.674 × 10⁻¹¹ N·m²/kg². The force on each body points toward the other, and by Newton’s third law the two forces are equal and opposite. When the Earth pulls a satellite with some force, the satellite pulls the Earth with exactly the same force; the Earth simply has so much mass that its acceleration is negligible.
The usual mistake is in r. For a satellite at height h above a planet of radius R, the distance in the law is measured from the planet’s center, $r = R + h$, not from the ground. Near the ground h is small compared with R, so $r \approx R$ and g hardly changes with height; putting h in place of r there would be wildly wrong (1 km up, it would make the pull about forty million times too strong). A few thousand kilometers up, h is no longer small beside R, so use $R + h$ in full: squaring the distance makes any error in it worse.
G is tiny, which is why we notice gravity only when one of the masses is a planet. The first measurement of it was Henry Cavendish’s in 1798, with a torsion balance designed by John Michell; Cavendish reported the Earth’s density, from which G follows. Small lead spheres on a light beam hang from a thin fiber, beside two large spheres; their pull twists the fiber, and a mirror on the beam throws a spot of light onto a distant scale, so a tiny twist moves the spot a long way. Moving the large spheres to the mirror-image position (A to B) shifts the spot’s rest point by ΔS. The fiber’s stiffness comes from the period T of the beam’s free swing, so neither needs to be measured directly. For spheres of mass M at center distance b from small spheres on a beam of half-length d, with the scale a distance L away,
$G = \dfrac{\pi^2b^2d\,\Delta S}{MT^2L}$
The far large sphere pulls each small sphere slightly the wrong way; this reduces the torque by a fraction $\beta = b^3/(b^2 + 4d^2)^{3/2}$, and the formula’s G is divided by $1 - \beta$. The Cavendish Bench also corrects for the pull growing as a small sphere swings closer, which softens the fiber. On its classroom balance, a model example rather than a real instrument (1.50 kg spheres, 38.3 g spheres, the scale 5.00 m away) the spot moves about 16 cm between A and B, and the swing takes about ten minutes. Knowing G, the Earth’s mass follows from $g$ and its radius: weighing the Earth.
Dividing the force by the small mass m gives the pull per kilogram, which depends only on the planet and the distance. That quantity has its own name.
The gravitational field, g
The acceleration a body would have if gravity alone acted on it is the gravitational field at that point, the force per unit mass:
$g = \dfrac{GM}{r^2}$, with $r = R + h$
At Earth’s surface, with $GM_E$ = 3.986 × 10¹⁴ m³/s² and $R_E$ = 6,371 km, $g_0$ = 9.82 m/s². (Real g varies with latitude, from about 9.78 m/s² at the equator to 9.83 at the poles, because the Earth spins and bulges at the equator; the familiar 9.81 m/s² is a typical value, not 9.82 less one spin correction. This sim leaves the spin and the bulge out.) Higher up, g falls as the inverse square of the distance from the center. It never reaches zero: gravity has no edge, and a body halfway to the Moon still feels the Earth.
A common trap is to assume g falls in a straight line with height, as if it were $g_0(1 - h/R)$. That guess has the wrong slope: for small heights the field falls about twice as fast, $g \approx g_0(1 - 2h/R)$, and farther up no straight line works. The correct ratio is $(R/r)^2$, so to find g at a height, square the ratio of the two radii. A second trap is believing there is “no gravity in space”. Low orbits are only a few hundred kilometers up, a small fraction of Earth’s radius, so g there is nearly as strong as on the ground.
Force and field describe gravity at one place. To compare places far apart, and to say what speed escapes the Earth entirely, energy is the better tool.
Gravitational energy and escape speed
Near the ground we write the potential energy as mgh, assuming g constant. Over larger distances the work done against the inverse-square force gives
$U = -\dfrac{GMm}{r}$
The zero is chosen at infinity, so U is negative everywhere else, rising toward zero as r grows. Only differences in U matter, and between two radii $\Delta U = GMm(1/r_1 - 1/r_2)$, which reduces to mgh when the heights are small. With gravity the only force, the total energy $E = \tfrac{1}{2}mv^2 - GMm/r$ is conserved.
A ball thrown upward slows as it climbs, because gravity keeps pulling it back the whole way. Gravity never switches off, but its pull weakens with distance, and a ball fast enough never runs out of speed. The slowest such launch arrives “at infinity” with zero speed and zero total energy. Setting E = 0 at the surface gives the escape speed:
$\tfrac{1}{2}mv_{esc}^2 - \dfrac{GMm}{R} = 0$ $v_{esc} = \sqrt{\dfrac{2GM}{R}}$
The mass m cancels, so a pebble and a spacecraft need the same speed. So does the direction. Kinetic energy depends on the speed, not on where the velocity points, so a ball launched sideways, straight up or at 45° escapes at the same speed, provided its path does not run into the ground on the way (this sim fires from a mountain above the air for that reason). It is tempting to think that straight up must need more, since the ball fights gravity head on, but the energy bill is the same.
Escape speed is also $\sqrt{2}$ times the speed of a circular orbit at the same distance, which the next section derives. Leaving out the 2 under the square root gives that circular speed instead, a common slip.
A ball launched below escape speed comes back, unless it is moving sideways fast enough to miss. That is an orbit.
An orbit is free fall
Return to Newton’s mountain. A ball fired sideways falls toward the Earth’s center, but in the time it falls, it moves sideways too. If the ground curves away by exactly as much as the ball falls, the ball stays at the same height forever: a circular orbit. The only force on it is gravity, and gravity supplies exactly the centripetal force a circle of radius r needs:
$\dfrac{GMm}{r^2} = \dfrac{mv^2}{r}$ $v_{orbit} = \sqrt{\dfrac{GM}{r}}$
Once around the circle, $2\pi r$, at that speed takes one period:
$T = \dfrac{2\pi r}{v_{orbit}} = 2\pi\sqrt{\dfrac{r^3}{GM}}$
As before, r is measured from the center, and the satellite’s own mass cancels: a wrench dropped beside a space station keeps pace with the station. Shots well below the circular speed land; from the 100 km mountain, a sideways ball must reach very nearly the circular speed to go all the way around without touching the ground.
Why, then, do astronauts float, when gravity on a low orbit is nearly as strong as on the ground? Because they and their spacecraft are falling together, with the same acceleration, so nothing presses on anything. Weight as we feel it is the push of a floor holding us up; in free fall nothing does. The explanation sometimes heard, that a “centrifugal force” cancels gravity, describes the same physics in a frame turning with the spacecraft (the Accelerating Frames sim explores it). In an inertial frame there is one force, gravity, and it is not balanced: it keeps turning the velocity.
The GPS satellite is slower than a low one, and takes far longer to go around. That comparison is the next idea.
Higher is slower, and geostationary orbits
Because $v = \sqrt{GM/r}$, a higher circular orbit is slower. That surprises many people, who reason that a satellite must go faster to stay up higher. The opposite is true: farther out gravity is weaker, and a weaker pull can bend only a slower path into a circle. The period grows even faster than the radius, because the satellite is both slower and has farther to go. Squaring the period formula gives Kepler’s third law for circles,
$T^2 = \dfrac{4\pi^2r^3}{GM}$
so four times the radius gives eight times the period. Read the other way, the law weighs a planet: time a moon around it and measure its distance, and $M = 4\pi^2r^3/GT^2$. This is how the masses of the planets are known.
One radius is special. A satellite whose period equals one turn of the Earth, moving eastward above the equator, keeps pace with the ground and appears to hang still in the sky: a geostationary orbit, where a dish can stay pointed at it. It must be over the equator, because every orbit’s plane passes through the Earth’s center; a satellite on a circle tilted away from the equator swings north and south each day, tracing a figure eight from the tilt alone (an eccentric orbit reshapes it).
One turn of the Earth is not 24 hours. The Earth turns once relative to the stars in a sidereal day, 23 h 56 min 4 s (86,164 s). The 24 h solar day is a little longer because the Earth also moves along its orbit, so it must turn slightly more than once to bring the Sun back overhead. A satellite with a 24 h period would lag behind the ground by about 1° each day, drifting slowly west. The Ground Track tab shows these tracks, and points a dish at a geostationary satellite from anywhere on Earth.
Speed and period describe an orbit’s motion. Its energy tells us what it costs to get there.
The energy of an orbit
On a circular orbit the speed is fixed by the radius, so the kinetic energy is too: $K = \tfrac{1}{2}mv^2 = GMm/2r$. Adding the potential energy gives the total:
$E = K + U = \dfrac{GMm}{2r} - \dfrac{GMm}{r} = -\dfrac{GMm}{2r}$
Three neat relations follow on every circle: $K = -E$, $U = 2E$, and $K = -\tfrac{1}{2}U$. The total energy is negative, as it must be for a satellite that cannot escape, and it rises toward zero for higher orbits. So a higher orbit has more total energy, even though it is slower.
Launching a satellite from the ground to a circular orbit means paying the difference between E in orbit and E on the launch pad, $-GMm/R$ (ignoring Earth’s spin). It is easy to count only the climb, $\Delta U$, or worse, mgh, as if the job were to lift the satellite like a crate onto a shelf. In fact most of the cost of reaching a low orbit is the speed, not the height.
A satellite that gets a little more than the circular speed, but less than escape speed, has an energy between these limits. It neither circles nor leaves.
Ellipses and equal areas
Fire Newton’s cannon a little faster than the circular speed. The ball climbs as it goes, slowing as it rises, reaches a highest point on the far side of the Earth, and falls back to the mountain, where it is fastest again: an ellipse. Kepler’s first law: orbits are ellipses with the attracting body at one focus. The nearest point is the perigee, the farthest the apogee (for orbits about the Sun, perihelion and aphelion). The long axis, $r_p + r_a$, is twice the semi-major axis a, and the eccentricity e measures how stretched the ellipse is, from 0 for a circle toward 1.
The energy of an elliptical orbit depends only on a, just as a circle’s depends only on r:
$a = \dfrac{r_p + r_a}{2}$ $E = -\dfrac{GMm}{2a}$
The sign of E sorts every orbit. Negative E is bound, a circle or an ellipse. E = 0 is the escape case, a parabola. Positive E is a hyperbola, leaving with speed to spare. A launch faster than circular but slower than escape therefore makes an ellipse, never a bigger circle or a slow spiral outward, and it comes back.
Kepler’s second law says a line from the focus to the planet sweeps out equal areas in equal times. Near perigee the line is short, so the satellite must move fast to sweep its share of area; near apogee it is long, and the satellite crawls. Equal areas are angular momentum in disguise: gravity pulls along the line to the center, so it exerts no torque about the center, and $L = mrv_\perp$ stays constant. At the two apsides the velocity is perpendicular to r, so
$r_pv_p = r_av_a$
A frequent mistake is to give the apogee the speed of a circle at that distance, $\sqrt{GM/r_a}$. At apogee the satellite is moving slower than a circle there would need, which is exactly why it falls back inward.
An ellipse that touches two circles connects them. That is how spacecraft move between orbits.
Moving between orbits
Suppose a satellite on a low circle must reach a high one. A forward burn, firing the engine along the direction of motion, adds speed and energy, and turns the circle into an ellipse whose perigee is the burn point; the far side of the orbit rises. If the burn is sized so that the apogee just reaches the high circle, the satellite coasts half an ellipse out to it. There it is moving slower than the high circle needs, so a second forward burn makes up the difference and the satellite stays. This is the Hohmann transfer, the textbook way between two circles.
The transfer ellipse touches both circles, so its semi-major axis is the average of their radii, and the coast between the burns is half its period:
$a = \dfrac{r_1 + r_2}{2}$ $t = \dfrac{T}{2} = \pi\sqrt{\dfrac{a^3}{GM}}$
Two slips are common here: using the full period, though only half the ellipse is flown, and taking a as the sum of the radii rather than their average. The surprise is in the speeds. Both burns point forward, yet the satellite ends on the high circle moving slower than it started, because higher circles are slower. The burns paid for energy, and the energy went into height.
Part 1 ends here. Part 2 goes inside the Earth, times orbits exactly, and counts the cost of every maneuver.
Part 2 · Going further
Inside the Earth: the shell theorem and the gravity tunnel
Newton’s shell theorem: outside a uniform spherical shell, the shell pulls as if all its mass were at its center; inside it, the pulls cancel exactly (the near side is closer, but there is more of the far side). A planet is a stack of shells, so at a distance r from its center only the mass M(r) inside r pulls:
$g(r) = \dfrac{GM(r)}{r^2}$
For a uniform planet $M(r) = M(r/R)^3$, so g falls in proportion to the distance from the center, $g = g_0r/R$, reaching zero at the center. The real Earth is not uniform: its iron core is about twice as dense as its rocky mantle. Going down a deep mine, you leave light rock behind while nearing the dense core, so g at first rises, peaking at the core–mantle boundary, about 2,900 km down. The sim uses PREM, the Preliminary Reference Earth Model built from seismic waves.
Now dig a straight, frictionless tunnel through a uniform Earth, along any chord, and drop a train in. At a distance x from the chord’s midpoint, the component of gravity along the tunnel is $-(GMm/R^3)x$, a force proportional to the displacement: simple harmonic motion, with $\omega = \sqrt{GM/R^3}$. The trip from one end to the other takes half a period,
$t = \pi\sqrt{\dfrac{R^3}{GM}}$
the same for every chord, short or long, because a shorter tunnel is less steep in exactly the right proportion. It is half the period of a satellite skimming the surface. Holding g at its surface value all the way down is wrong for a uniform planet, where g shrinks toward the middle; for the real Earth, whose g hardly falls through the mantle, it comes close, and the layered PREM Earth makes the trip several minutes shorter than the uniform one (Klotz, 2015).
Vis-viva and Kepler’s equation
Energy conservation with $E = -GMm/2a$ gives the speed anywhere on any orbit, from the distance r and the semi-major axis alone. Writing $\tfrac{1}{2}mv^2 - GMm/r = -GMm/2a$ and solving,
$v^2 = GM\left(\dfrac{2}{r} - \dfrac{1}{a}\right)$
This is the vis-viva equation (“living force”, the old name for kinetic energy). On a circle r = a and it gives $v^2 = GM/r$. Orbits with the same a share their energy and period, however stretched; the mistake to avoid is a circle’s speed at a point of an ellipse.
Kepler’s equation gives the time to any point. Measure the position by the true anomaly θ, the angle at the focus from perigee. Define an auxiliary angle ψ, the eccentric anomaly (astrodynamics texts write E, but E is the energy here), by
$\tan\dfrac{\psi}{2} = \sqrt{\dfrac{1-e}{1+e}}\tan\dfrac{\theta}{2}$
Then the time t since perigee satisfies
$\dfrac{2\pi t}{T} = \psi - e\sin\psi$
The left side, the mean anomaly, grows steadily with time. Going from a position to a time needs only arithmetic; the reverse has no closed form, and the sim solves it numerically.
Eccentric orbits use this lopsided timing. A Molniya orbit, with a 12-hour period, its apogee high over the north and a tilt of 63.4° (where the Earth’s bulge leaves the apogee in place), spends most of each lap where geostationary satellites sit too low on the horizon to serve.
Δv budgets
A rocket’s fuel buys changes of velocity, so mission planners count the Δv of each burn and add them up. Δv is the currency because the rocket equation (Momentum & Collisions) turns it directly into a mass of fuel. Speeds are never added; burns are. For a Hohmann transfer from a circle $r_1$ to a circle $r_2$, each burn is the difference between the speed the spacecraft has and the speed the next orbit needs, from vis-viva with $a = (r_1 + r_2)/2$:
$\Delta v_1 = \sqrt{\dfrac{GM}{r_1}}\left(\sqrt{\dfrac{2r_2}{r_1 + r_2}} - 1\right)$ $\Delta v_2 = \sqrt{\dfrac{GM}{r_2}}\left(1 - \sqrt{\dfrac{2r_1}{r_1 + r_2}}\right)$
The difference of the two circular speeds is not the cost: it is what the spacecraft loses on the way up, not what it spends. Two burns are not always cheapest: a bi-elliptic transfer burns out to a far apogee beyond the target, raises the perigee there cheaply, then falls back and brakes into the target circle. For a target more than 11.94 times the starting radius, a far enough apogee beats Hohmann; beyond 15.58 times, any apogee past the target does. The saving is small and the trip much longer, so it is rarely flown.
A velocity can change direction as well as size. To tilt an orbit by Δi, the burn must turn the velocity, and if the speed changes too, from $v_1$ to $v_2$, the law of cosines gives the cost:
$\Delta v = \sqrt{v_1^2 + v_2^2 - 2v_1v_2\cos\Delta i}$
At constant speed it is $2v\sin(\Delta i/2)$: a 60° turn costs as much as the speed itself. So planes are changed where the spacecraft is slowest, at apogee, combined with a burn needed anyway.
Catching another spacecraft in the same orbit is called phasing, and it runs against intuition. To catch a station ahead, burn backward: the far side of your orbit drops, the new orbit is shorter, and after one lap you return to the burn point early, just as the station arrives. A forward burn would send you higher and slower, and you would fall behind.
Leaving the Earth, and launching from it
A probe bound for another planet must leave the Earth with speed to spare, the hyperbolic excess speed $v_\infty$ it keeps far away. Energy conservation between a parking orbit of radius r and “infinity” gives the speed it needs at r:
$\tfrac{1}{2}v_p^2 - \dfrac{GM}{r} = \tfrac{1}{2}v_\infty^2$ $v_p = \sqrt{v_{esc}^2 + v_\infty^2}$
and the burn is $v_p$ minus the circular speed. Escaping first and then adding $v_\infty$ is the tempting mistake: energies add, not speeds, and a burn made where the probe is already fast adds the most energy, $v\Delta v + \tfrac{1}{2}\Delta v^2$ per kilogram. This is the Oberth effect, the reason departure burns are made at perigee, close to the planet.
Before any of that, the rocket must reach orbit, and the Earth helps. The ground already moves eastward with the Earth’s spin, at $\omega R\cos\lambda$ at latitude λ, where $\omega = 2\pi/T_{sid}$. A rocket launched due east starts with that speed and needs that much less; one launched west must first cancel it. So launch sites sit as near the equator as they can, and most launches head east. Real rockets need about 9.4 km/s of Δv to reach low orbit, since gravity and air take a share during the climb.
Drag and orbital decay
At a few hundred kilometers the air is billions of times thinner than at sea level, but a satellite moving at nearly 8 km/s still meets enough of it to feel drag. Drag acts against the motion and takes energy away, so you might expect the satellite to slow down. It does the opposite. On a near-circular orbit $E = -K$, so losing energy raises the kinetic energy: the satellite spirals down to a lower, faster orbit. Gravity does twice as much work on the way down as drag takes. This is the drag paradox: a decaying satellite speeds up as it falls, until the thicker air low down brings it in. A compact, dense satellite (a large ballistic coefficient $B = m/C_DA$) lasts longest, and an active Sun, which puffs up the upper air, shortens every lifetime.
Tides
The Moon pulls harder on the side of the Earth facing it than on the Earth’s center, and harder on the center than on the far side. Relative to the center, the near side is pulled toward the Moon and the far side left behind: the Earth is stretched along the Earth–Moon line into two bulges of ocean, and it turns under them, giving two high tides a day. That is the equilibrium sketch, not a forecast: the ocean basins, the coastlines and the Moon’s tilt make real tides unequal, and some coasts get only one high tide a day. The stretch is the difference of the pull across the Earth’s radius, which for a distant body is approximately
$\Delta a \approx \dfrac{2GMR}{d^3}$
One power of d more than the pull itself. So the Sun, whose pull on the Earth is far stronger than the Moon’s, raises the smaller tide, because it is so much farther away. Taking the ratio of the pulls, $1/d^2$, instead of the stretches is the usual slip. Lined up, at new and full moon, the two tides add (spring tides); at the quarter moons they partly cancel (neap tides). The same stretch on the Moon long ago locked its spin to its orbit.
Reference
Common mistakes
- r from the surface. Every formula here measures r from the center: $r = R + h$.
- “No gravity in orbit.” Low-orbit g is nearly its surface value; astronauts float because they fall.
- Higher means faster. Higher orbits are slower and take much longer, though they have more energy.
- Escape needs aiming up. Energy has no direction: any launch that clears the ground escapes at the same speed.
- Faster than circular means leaving. Below escape speed the path is an ellipse that comes back.
- A circle’s speed at apogee. On an ellipse use $r_pv_p = r_av_a$ or vis-viva.
- The full period for a transfer. A Hohmann transfer flies half an ellipse, with $a = (r_1 + r_2)/2$.
- Adding $v_\infty$ to escape speed. Energies add, not speeds.
- A 24-hour day for geostationary orbit. The Earth turns once in a sidereal day, 23 h 56 min 4 s.
Equation sheet
| Part 1 | |
|---|---|
| gravity | $F = \dfrac{GMm}{r^2}$; $g = \dfrac{GM}{r^2}$, $r = R + h$ |
| energy | $U = -\dfrac{GMm}{r}$; $v_{esc} = \sqrt{\dfrac{2GM}{R}}$ |
| circles | $v_{orbit} = \sqrt{\dfrac{GM}{r}}$; $T = 2\pi\sqrt{\dfrac{r^3}{GM}}$ |
| weighing | $M = \dfrac{4\pi^2r^3}{GT^2}$ |
| orbit energy | $E = -\dfrac{GMm}{2r}$; $E = -\dfrac{GMm}{2a}$ |
| ellipses | $r_pv_p = r_av_a$; $a = \dfrac{r_p + r_a}{2}$ |
| Cavendish | $G = \dfrac{\pi^2b^2d\,\Delta S}{MT^2L}$, then divide by $1 - \beta$ |
| Part 2 | |
|---|---|
| inside | $g = g_0\dfrac{r}{R}$ (uniform); $t = \pi\sqrt{\dfrac{R^3}{GM}}$ |
| vis-viva | $v^2 = GM\left(\dfrac{2}{r} - \dfrac{1}{a}\right)$ |
| Kepler | $\dfrac{2\pi t}{T} = \psi - e\sin\psi$; $\tan\dfrac{\psi}{2} = \sqrt{\dfrac{1-e}{1+e}}\tan\dfrac{\theta}{2}$ |
| Hohmann | $\Delta v_1 = \sqrt{\dfrac{GM}{r_1}}\left(\sqrt{\dfrac{2r_2}{r_1 + r_2}} - 1\right)$ |
| $\Delta v_2 = \sqrt{\dfrac{GM}{r_2}}\left(1 - \sqrt{\dfrac{2r_1}{r_1 + r_2}}\right)$ | |
| plane change | $\Delta v = \sqrt{v_1^2 + v_2^2 - 2v_1v_2\cos\Delta i}$ |
| departure | $v_p = \sqrt{v_{esc}^2 + v_\infty^2}$; Oberth: $\Delta E/m = v\Delta v + \tfrac{1}{2}\Delta v^2$ |
| spin, tides | $v_{ground} = \omega R\cos\lambda$; $\Delta a \approx \dfrac{2GMR}{d^3}$ |
Symbols: G = 6.674 × 10⁻¹¹ N·m²/kg² · M, m the two masses · r distance from the center · R the planet’s radius, h height · $g_0$ the surface field · T period · a semi-major axis · $r_p$, $r_a$, $v_p$, $v_a$ perigee and apogee distances and speeds · e eccentricity · θ true anomaly (from perigee) · ψ eccentric anomaly · Δi change of tilt · $v_\infty$ speed left far away · ω the Earth’s spin, λ latitude · d distance to the tide-raising body · b, d, L, ΔS, β the Cavendish balance.
Further reading
OpenStax, University Physics Volume 1, chapter 13, Gravitation, whose order and notation Part 1 follows; for orbits, Curtis, Orbital Mechanics for Engineering Students, and Vallado, Fundamentals of Astrodynamics and Applications. The Field Manual lists every source and what the model leaves out.
Gravitation & Satellite Orbits v1.3 · © 2026 Dr. David Meier · CC BY-NC-SA 4.0 · glidepathscience.org