Aerocapture — using atmosphere to decelerate into orbit: Question bank

Before we start, one anchor so no symbol sneaks in undefined:

Recall The symbols this page assumes (reveal if unsure)

$ε$ ::: specific orbital energy (energy per kilogram), $ε = v^{2} /2 - μ / r$ . Its sign decides the orbit type.
$μ = GM$ ::: the planet's gravitational parameter (mass $\times$ Newton's constant).
$v_{\infty}$ ::: "excess" speed — the speed left over infinitely far from the planet, the leftover of a hyperbolic arrival.
$β = m / (C_{D} A)$ ::: ballistic coefficient — mass divided by drag-area. Low $β$ = light-and-blunt.
$ρ (h) = ρ_{0} e^{- h / H}$ ::: air density falling exponentially with height, $H$ the scale height.
$a_{D} = ρ v^{2} / (2 β)$ ::: drag deceleration per kilogram.

True or false — justify

The atmosphere removes energy, so a captured orbit always has less energy than the arrival hyperbola.

True. Drag only ever takes energy out (

ε

drops from positive to negative); it can never add energy, so

ε_{o u t} < ε_{in}

always.

A spacecraft is "captured" the moment its speed drops below the local circular-orbit speed.

False. Capture is decided by the sign of $ε = v^{2} /2 - μ / r$ , i.e. speed below the local escape speed

2 μ / r

, not the circular speed. Between circular and escape speed you are still bound.

Aerocapture and aerobraking obey different drag laws.

False. Both use the same drag

F_{D} = \frac{1}{2} ρ v^{2} C_{D} A

. They differ only in how much and how many times — aerocapture: one deep pass; aerobraking: hundreds of shallow ones on an already-bound orbit.

If a probe skips back out of the atmosphere still on a hyperbola, aerocapture has failed.

True. Skipping out with

ε > 0

means not enough energy was shed; the craft escapes and is lost. Success requires exiting with

ε < 0

Lowering periapsis always increases the chance of successful capture.

False. Because density is exponential, dropping a few km can overshoot into runaway drag — you over-decelerate (crash) or overheat. Deeper helps only up to the corridor's lower edge, then it kills you.

A parabolic arrival (

ε = 0

) needs zero drag to be captured.

False. A parabola is the marginal escape case; it is not bound. You still need to remove a finite amount of energy to push

ε

strictly below zero.

The

Δ v

that drag supplies is genuinely free in the fuel sense.

True.

Δ v_{a er o} = v_{in} - v_{o u t}

is speed killed by air, so by the rocket equation it is propellant you never had to launch — the entire economic point.

Two identical-mass probes with different frontal areas will brake identically at the same altitude.

False. Braking scales as

1/ β = C_{D} A / m

. Larger area (lower

β

) means stronger deceleration for the same air, so they behave very differently.

Peak heating and peak deceleration happen at exactly the same instant.

False (generally). Deceleration peaks near where

ρ v^{2}

maximises; stagnation heating scales as

ρ^{1/2} v^{3}

, a different combination, so its peak occurs at a different point along the trajectory.

Spot the error

"Drag is a force and force changes momentum, so we should decide capture by momentum bookkeeping."

Error: momentum is not conserved on a curved gravitational pass and its sign says nothing about orbit type. Capture is an energy question — test the sign of

ε

, never momentum.

"Entering faster is better because

F_{D} \propto v^{2}

gives more braking."

Error: yes drag rises with

v^{2}

, but the energy you must remove also rises with

v^{2}

, and heating rises with

v^{3}

. Faster arrival is harder, because heating (the cubic term) is the true limiting constraint.

"A high ballistic coefficient is safer because the craft is dense and sturdy."

Error: high

β

means weak braking high up, so the craft plunges deeper into dense air before slowing — hotter and riskier. Low-

β

blunt shields brake high and cool.

"Aim for the densest part of the corridor to guarantee enough drag."

Error: the densest edge is the dangerous edge (runaway drag, heating spike). You target the middle of the corridor and trim with lift/bank angle, not by diving deeper.

"Since the craft is only in the air a short time, we can ignore how energy is lost and just use

Δ v = v_{in} - v_{o u t}

Partly-error:

Δ v_{a er o} = v_{in} - v_{o u t}

is a valid bookkeeping of the speed killed at periapsis, but what causes it is the drag integral over the pass — you cannot design the pass without

a_{D} = ρ v^{2} / (2 β)

and the density profile.

"Because

ε = - μ / (2 a)

, a more negative energy means a bigger orbit."

Error: more negative

ε

means smaller semi-major axis

a

(tighter orbit). Over-braking pushes

ε

too negative and shrinks the orbit toward the surface.

Why questions

Why does the exponential atmosphere make periapsis altitude a razor's-edge knob?

Because

ρ

changes by a factor

e

every scale height

H

(a few km), a small altitude error multiplies the drag hugely — so the tolerable band of altitudes (the corridor) is narrow.

Why can't you simply do one enormous drag pass with any old heat shield?

Stagnation heating scales as

ρ^{1/2} v^{3}

; a single hyperbolic-speed pass concentrates all the braking (hence heat) into minutes, demanding a robust thermal protection system — this is why aerocapture is harder than gentle aerobraking.

Why is aerocapture worth the risk versus a propulsive insertion burn?

By Tsiolkovsky, saving

Δ v

saves a mass fraction

1 - e^{- Δ v / v_{e}}

; killing ~0.8 km/s with

v_{e} \approx 3

km/s saves nearly a quarter of the arrival mass in propellant not carried.

Why must there be both an upper and a lower boundary to the entry corridor?

Too high (upper edge): insufficient drag → you skip back to escape. Too low (lower edge): excessive drag/heating → crash or burn up. Survival requires threading between the two.

Why does vis-viva guarantee the exit speed is fixed once you fix exit altitude and target orbit?

Because

ε = v^{2} /2 - μ / r

ties speed to radius for a given energy; choosing the target

ε_{o u t}

and the exit

r

pins down

v_{o u t}

uniquely.

Why does lift (bank-angle) steering beat depth-only control?

Lift lets you adjust the effective time and altitude in the atmosphere continuously and reversibly, whereas depth is set at atmospheric entry and cannot be undone once you are committed to a periapsis.

Edge cases

v_{\infty} = 0

(arriving with exactly zero excess speed), what does aerocapture need to do?

Then

ε_{in} = 0

(parabolic, marginal). You still need a small drag pass to push

ε

below zero, but the required braking is minimal — this is the easiest limit.

What happens as

β \to \infty

(extremely dense, tiny frontal area)?

a_{D} = ρ v^{2} / (2 β) \to 0

: the craft barely feels the air at any survivable altitude, so aerocapture becomes impossible and you must burn propellant instead.

What happens as

β \to 0

(huge area, tiny mass — a "feather")?

Braking becomes enormous at very high, thin altitudes; the craft can be captured gently and cool, but structurally such a giant-area shield is hard to build — the practical limit favours low but finite

β

At the exact instant

ε

passes through zero during the pass, what is the orbit?

Instantaneously parabolic — the boundary between escape and capture. If drag stops there the craft is on the knife-edge; it must keep shedding energy to stay bound.

If drag removes too much energy so

ε

becomes very negative, what is the outcome?

The semi-major axis

a = - μ / (2 ε)

shrinks so far that periapsis (or the whole orbit) dips into the surface — over-capture becomes a crash, the lower-corridor failure.

Can aerocapture ever turn a bound ellipse back into a hyperbola?

No. Drag only removes energy, so it can never raise

ε

from negative back to positive — the reverse direction requires a burn, which is why aerobraking a captured orbit is safe from accidental escape.

Recall One-line self-test before you leave

The single question that resolves most traps on this page ::: "Is $ε$ positive or negative — and by how much?" Sign decides captured-vs-escape; magnitude decides orbit size and whether you over-braked.