Appendix H: Launch Vehicle and Engine Reference

This appendix is a one-stop ballpark reference for the launch vehicles and rocket engines that recur throughout the book — the hardware behind the physics of Part III (Propulsion) and Part V (Mission Design and Operations). When a worked example says "Falcon 9's first stage" or "a staged-combustion engine like the RS-25," this is where you can find the rough numbers in one place, and see how one vehicle or engine stacks up against another. It is a table to build intuition with, not a table to design with.

⚠️ Read this before you trust a single number below. Every performance and specification figure in this appendix is approximate and Tier 2 (widely reported, no single authoritative source) — or Tier 3 (aspirational/round) where noted. These numbers are soft for real reasons, not sloppy ones: a vehicle's payload depends on its exact version and block, the number of strap-on boosters, the target orbit and inclination, the launch site, and — critically — whether the stage is recovered or thrown away. Engines are uprated across versions, run at different throttle or power levels, and are quoted at either sea level or vacuum. Treat everything here as planning intuition, good to about one or two significant figures — never as a value to design, price, or fly a mission with. Where a figure is especially soft or officially unpublished, the notes say so. For engineering work, use the operator's current payload user's guide and the engine maker's data sheet.

How to read these tables

A few conventions make the tables compact:

• Payload to LEO (low Earth orbit, roughly 200–500 km) is the headline "how much can it lift" number. Payload to GTO (geostationary transfer orbit) is much smaller because the vehicle must supply most of the delta-v to nearly reach GEO — see Appendix G's delta-v map and Chapter 9 for why these two numbers differ by roughly a factor of three.
• Ranges (e.g., "~11–27") span a family's smallest to largest configuration — usually the number of solid boosters. A single "~" number is a representative maximum for the common version.
• Reuse cuts payload. A partially reusable vehicle lifts less when it recovers its booster than when it expends it, because the propellant saved for landing is propellant not spent on the payload. Falcon 9 is the clearest example: roughly 17 t to LEO recovering the booster versus about 22.8 t expending it. Numbers here lean toward the higher (often expendable) figure unless noted.
• Propellant shorthand: kerolox = RP-1 (refined kerosene) + liquid oxygen; methalox = liquid methane + LOX; hydrolox = liquid hydrogen + LOX; hypergolic = self-igniting storables (typically UDMH + nitrogen tetroxide); SRB = solid rocket booster.

H.1 Launch vehicles

Fifteen vehicles that between them fly most of the world's payloads, plus a few in development that will reshape the columns. Tonnages are metric tons (t); all are approximate (Tier 2), version-dependent, and tied to a specific orbit and reuse mode.

Notes on the launch-vehicle table.

• The config swing is real. Vehicles with strap-on solids (Atlas V, Vulcan, Ariane 6, LVM3, SLS) change capability dramatically with booster count. Vulcan spans roughly 11 t (no boosters) to 27 t to LEO (six boosters, the "VC6" configuration); Atlas V's "401" through "551" designations encode the same idea. Always ask which version a number refers to.
• GTO is not everyone's target. For SLS, the meaningful number is not GTO but throw to the Moon: on the order of ~27 t to trans-lunar injection (Block 1), with later blocks planned toward ~130 t to LEO — those later-block figures are design goals, so treat them as especially soft. Electron is a dedicated small-lift vehicle; it reaches beyond LEO only with small payloads on a kick stage, so its GTO cell is left blank rather than forced.
• Starship is a moving target by definition. As a vehicle still in development, its payload is a design goal (~100–150 t to LEO, fully reusable) rather than a demonstrated capability, and it reaches high-energy orbits through on-orbit refueling rather than a quoted GTO number. Read its row as intent, not record.
• "Operational (retiring)" marks vehicles still flying out manifests while their successors ramp up: Atlas V (giving way to Vulcan) and Proton (to Angara). New Glenn is newly operational, so its numbers are Tier 2 and likely to firm up with flight experience.

H.2 Rocket engines

Ten engines that anchor the propulsion chapters. Specific impulse ($I_{sp}$) is quoted in seconds at sea level and in vacuum; thrust is in kilonewtons (kN), labeled as sea-level (SL) or vacuum (vac) because the two differ by 10–30%. A dash means the value does not apply (an upper-stage engine has no meaningful sea-level rating) or is not publicly established. All figures are Tier 2 and version-dependent.

Notes on the engine table.

• Cycle sets the character. The "cycle" column is the single most explanatory fact about an engine — it largely determines efficiency, chamber pressure, thrust, and complexity. Gas-generator engines (Merlin, F-1, Vulcain 2) burn a little propellant to drive the pumps and dump it overboard, trading some $I_{sp}$ for simplicity. Staged-combustion engines (RS-25, RD-180, BE-4, Raptor) feed that turbine exhaust back into the chamber for higher efficiency and pressure, at the cost of much greater complexity. Expander engines (RL10, Vinci) use the fuel itself, warmed by the chamber, to spin the pumps — elegant and efficient, but practical only at the modest thrust of upper stages. Chapter 17 explains each cycle and why an engine's designers chose it.
• Sea level versus vacuum. An engine always delivers higher $I_{sp}$ and thrust in vacuum, where no atmospheric back-pressure fights the nozzle. The gap is largest for high-expansion hydrolox engines: the RS-25 climbs from ~366 s at sea level to ~452 s in vacuum. Upper-stage engines (RL10, Vinci) are built only for vacuum and have no sea-level rating.
• Power levels and versions. The RS-25 is quoted at its "100% power level" (~2,090 kN vacuum); it is certified to throttle up to about 109%, roughly ~2,280 kN. The Raptor row is Raptor 2; the family has climbed from roughly 1,800 kN (Raptor 1) toward ~2,750 kN (Raptor 3), a textbook case of version dependence. The vacuum-optimized Raptor ("RVac") reaches about ~380 s $I_{sp}$. Merlin likewise flies a vacuum-optimized variant on Falcon's second stage with $I_{sp}$ near ~348 s.
• The soft ones, flagged. Blue Origin has not published official $I_{sp}$ for the BE-4; the ~310–340 s figures here are third-party estimates, and its thrust has been publicly uprated from about 2,400 kN toward ~2,800 kN, so both BE-4 cells are the least certain in the table. Hydrolox upper-stage engines (RL10, Vinci) and the RS-25 hold the efficiency crown precisely because hydrogen's low molecular weight maximizes exhaust velocity — the physics in Chapter 17.

H.3 A word on cost per kilogram

Of every number attached to a rocket, cost per kilogram to orbit is the softest and most abused. Quoted prices are not the same as internal costs; a rideshare seat differs wildly from a dedicated launch; and the per-kilogram figure moves with the target orbit and how full the rocket is. Treat the following as order-of-magnitude only — Tier 2 for today, and frankly Tier 3 (speculative) for tomorrow:

• Expendable launch, historically: very roughly $10,000–$20,000 per kg to LEO (Tier 2). The Space Shuttle, fully accounted, was far higher — often cited on the order of tens of thousands of dollars per kilogram.
• Partial reuse (Falcon 9): advertised prices have fallen to roughly $2,000–$3,000 per kg (Tier 2), with internal cost reportedly lower still. This roughly order-of-magnitude drop is the reason reusability is a theme of this book, not a footnote.
• Full reuse (Starship goal): aspirations run one to two orders of magnitude lower again — figures like under $1,000 per kg*, or the frequently repeated *~$100 per kg or even ~$10 per kg, are aspirational and unproven (Tier 3), not demonstrated performance. Quote them as ambition, never as fact.

The honest one-sentence version: reusability has already cut launch price by roughly an order of magnitude and may cut it by another, but the exact numbers depend on business decisions and flight rates that no equation in this book can pin down. Chapter 38 tells that story.

H.4 Where the physics and the stories behind these numbers live

These tables are the what. The rest of the book is the why:

• Chapter 17 — engine cycles (gas-generator, staged combustion, full-flow, expander, electric-pump) and how a cycle sets an engine's $I_{sp}$, thrust, chamber pressure, and complexity. Every row of the engine table is an instance of a cycle explained there.
• Chapter 30 — launch-vehicle families and how you actually select one: matching payload, orbit, fairing volume, schedule, and price to a mission, using tables like these as the first cut.
• Chapter 38 — the reusability revolution (Falcon 9, Starship, Raptor) that is rewriting the payload, reuse, and cost columns even as you read them.

A note on precision: the figures here are teaching values — rounded, widely reported, and true to maybe one or two significant figures. If two sources disagree on a vehicle's payload or an engine's thrust, they are usually quoting different versions, configurations, power levels, or orbits, not contradicting each other. When a chapter needs a specific, defensible number, it cites its source and says so. For anything you intend to build, price, or fly, go to the primary data sheet — and expect it to have changed since this was written.