Aerobraking — gradual orbit lowering using atmospheric drag

WHY do we aerobrake at all?

When a probe arrives at Mars (or Venus, or returns to Earth) it is usually captured into a big, stretched-out elliptical orbit. To do useful science we often want a small, near-circular orbit. Lowering that orbit chemically requires a large velocity change $\Delta v$, and by the rocket equation every $\Delta v$ costs an exponential amount of propellant:

$\frac{m_{\text{fuel}}}{m_{\text{final}}} = e^{\Delta v / v_e} - 1$

WHAT is actually happening geometrically?

An elliptical orbit has two key points:

• Periapsis ($r_p$): closest to the planet — highest speed.
• Apoapsis ($r_a$): farthest — lowest speed.

Drag only matters where the air is dense — i.e. near periapsis. A brake pulse at periapsis lowers the speed there, which drops the apoapsis on the opposite side.

HOW does a speed loss at periapsis lower the apoapsis? (Derivation)

Step 1 — Energy of an orbit. Total specific mechanical energy (per unit mass): $\varepsilon = \frac{v^2}{2} - \frac{\mu}{r}, \qquad \mu = GM$ Why this step? Energy combines kinetic and gravitational potential; it is conserved in a pure gravity field, so it labels the whole orbit with one number.

Step 2 — Energy fixes the semi-major axis. For any bound orbit, $\varepsilon = -\frac{\mu}{2a}$ Why? Evaluate $\varepsilon$ at periapsis and apoapsis using $r_p = a(1-e)$, $r_a=a(1+e)$ and vis-viva; the $r$-dependence cancels leaving only $a$. So less energy (more negative) ⇒ smaller $a$ ⇒ smaller orbit.

Step 3 — Vis-viva (speed anywhere). From $\varepsilon = v^2/2 - \mu/r = -\mu/2a$, solve for $v$: $\boxed{v^2 = \mu\!\left(\frac{2}{r} - \frac{1}{a}\right)}$

Step 4 — Drag acts at periapsis, where $r=r_p$ stays fixed. A small drag $\Delta v < 0$ at periapsis changes energy by $\Delta\varepsilon = v_p\,\Delta v \quad(\text{from } d(\tfrac12 v^2)=v\,dv)$ Since $\varepsilon = -\mu/2a$: $\Delta a = \frac{2a^2}{\mu}\,\Delta\varepsilon = \frac{2a^2 v_p}{\mu}\,\Delta v$ Why this step? It links the tiny speed loss to how much the orbit shrinks. Because $v_p$ (periapsis speed) is large, each pass is efficient.

Step 5 — What happens to apoapsis? Since $r_p$ is unchanged (the brake happens at $r_p$) and $2a = r_p + r_a$: $\Delta r_a = 2\,\Delta a = \frac{4a^2 v_p}{\mu}\,\Delta v$ So a negative $\Delta v$ at periapsis pulls the apoapsis down by roughly four times the change in $a$. The near side ($r_p$) barely moves — exactly what we want: gently lower the far side pass after pass.

HOW much does one pass slow you? (The drag part)

Deceleration from atmospheric drag: $a_D = \frac{1}{2}\,\frac{\rho\, v^2 C_D A}{m}$ Why this form? Momentum swept out per second $=\rho A v \cdot v$; $C_D$ accounts for shape; divide by mass for acceleration. Density $\rho$ falls off exponentially with altitude $h$: $\rho = \rho_0 e^{-h/H}$ ($H$ = scale height). So drag is fiercely sensitive to periapsis altitude — the whole art of aerobraking is nudging $r_p$ to keep heating and $\Delta v$ per pass safely small.

Forecast-then-Verify

Common mistakes (Steel-manned)

Flashcards

Connections

• Vis-Viva Equation — the speed–radius–$a$ backbone used here.
• Orbital Energy and Semi-major Axis — why $\varepsilon=-\mu/2a$.
• Tsiolkovsky Rocket Equation — the exponential fuel cost we avoid.
• Aerocapture — the single-pass cousin.
• Atmospheric Drag and Scale Height — where $\rho=\rho_0e^{-h/H}$ comes from.
• Hohmann Transfer — the all-propulsive alternative for changing orbit size.

Concept Map

Hinglish (regional understanding)