Tsiolkovsky Rocket Equation & Delta-V Budget
A single equation, derived in 1903 by a Russian schoolteacher, governs every launch vehicle ever built. The rocket equation tells you exactly how much velocity change (Δv) you can wring from a given mass of propellant — and it turns out the answer is brutally unforgiving.
1. Deriving the Equation
Consider a rocket of total mass m in empty space. In a small time dt, it ejects exhaust mass dm at speed v_e (relative to the rocket). By conservation of momentum:
This is the Tsiolkovsky rocket equation. The result is logarithmic, which has a profound implication: to double Δv you don't just double the fuel — you need to square the mass ratio.
2. Specific Impulse (I_sp)
The exhaust velocity v_e is usually quoted as specific impulse I_sp, measured in seconds:
I_sp is the number of seconds one kilogram of propellant can produce one Newton of thrust — a fuel-independent figure of merit. Higher I_sp means more Δv per kilogram of propellant.
| Propellant Combination | I_sp (vac, s) | v_e (km/s) | Application |
|---|---|---|---|
| Solid rocket booster | 250–280 | 2.45–2.75 | SRB boosters |
| RP-1 / LOX (kerosene) | 350–358 | 3.43–3.51 | Falcon 9, Atlas V |
| LH₂ / LOX (cryogenic) | 440–460 | 4.31–4.51 | Space Shuttle main, Ariane 5 upper |
| Methane / LOX (Methalox) | 363–380 | 3.56–3.73 | Raptor (Starship), BE-4 (New Glenn) |
| Hydrazine (monoprop) | 220–235 | 2.16–2.30 | Satellite thrusters |
| Ion thruster (xenon) | 1,500–10,000 | 14.7–98 | Deep space probes, station-keeping |
3. The Tyranny of the Rocket Equation
The equation's logarithm makes large Δv missions exponentially expensive in propellant mass. Let's compute the propellant fraction f_p for various Δv values at I_sp = 350 s (v_e = 3.43 km/s):
When 98% of your launch mass must be propellant, only 2% is left for engines, structure, and payload. This is not a design choice — it is a mathematical law. The only escapes are: higher I_sp, multi-staging, or not needing as much Δv (e.g. in-space propulsion).
4. Multi-Stage Rockets
Once a stage burns out, its empty tanks and engines are dead weight. Discarding them restores a favourable mass ratio for the remaining stages. The total Δv of an N-stage rocket is the sum:
Example — two-stage comparison vs single stage to LEO (Δv needed = 9.4 km/s, I_sp = 350 s both stages):
- Single stage: Mass ratio = e^(9400/3430) = 15.5. Payload fraction = 1/15.5 − structure ≈ 0% (impossible in practice)
- Two stages equal mass ratio: Each stage delivers Δv = 4700 m/s with R = √15.5 = 3.94. Combined payload fraction ≈ 3–5%
- Three stages: Payload fraction ≈ 5–8%; used by Saturn V (3 stages to TLI)
SpaceX Falcon 9 achieves ~4% payload fraction to LEO. Starship/Super Heavy targets ~8% to LEO through propellant transfer and full reusability.
5. Delta-V Budgets
Mission designers build a Δv budget — a ledger of every manoeuvre needed:
| Manoeuvre | Typical Δv (m/s) | Notes |
|---|---|---|
| Launch to LEO (200 km) | 9,200–9,800 | includes gravity & drag losses |
| LEO → GTO Hohmann burn 1 | 2,440 | raises apogee to 35,786 km |
| GTO → GEO circularisation | 1,470 | raises perigee, plane change |
| LEO → Lunar orbit | 3,130 | TLI + LOI |
| Lunar landing | ~1,900 | descent from 15 km |
| Mars orbit insertion | ~800 | aerobraking reduces this |
| Falcon 9 booster return | ~1,350 | flip, boost-back, entry, landing |
6. Gravity & Drag Losses
The ideal Δv from the rocket equation applies in vacuum with no gravity. Actual launches incur losses:
- Gravity losses (~1,000–1,500 m/s): Thrust wasted fighting gravity during the vertical climb. Minimized by pitching over early and flying a gravity-turn trajectory (thrust vector aligned with velocity).
- Drag losses (~50–150 m/s): Aerodynamic drag mainly in the dense lower atmosphere. Minimized by small cross-section and a "max-Q throttle-down" as dynamic pressure peaks at ~12–14 km altitude.
- Steering losses (<50 m/s): Thrust slightly off the velocity vector for guidance corrections.
Launching east exploits Earth's rotation: equatorial surface speed ≈ 465 m/s. Florida launches (28.5°N) get ≈ 409 m/s for free. Kourou (5.2°N) gets ≈ 463 m/s — a significant advantage for GEO missions.
7. Beyond Chemical Rockets
The rocket equation is inescapable but the parameters aren't fixed. Future propulsion approaches:
- Nuclear thermal (NTR): I_sp 800–1,000 s. Heat H₂ propellant with a fission reactor. Doubles efficiency vs LH₂/LOX. NERVA demonstrated in the 1960s. Under development again for lunar logistics.
- Nuclear electric (ion): I_sp 2,000–10,000 s. Solar or fission power drives ion thrusters. Very low thrust (millinewtons) but extraordinary efficiency for deep-space probes. Hayabusa2, Dawn, BepiColombo all use ion drives.
- Solar sail: Zero propellant. Radiation pressure provides continuous tiny thrust. IKAROS (JAXA 2010), LightSail 2 (Planetary Society 2019) demonstrated. Best for inner-solar-system missions with time to spare.
- Laser launch / Lightcraft: Ground-based laser heats propellant or ablates the vehicle. No propellant mass needed if power is remote. Demonstrated at gram-scale; scaling is the challenge.