# Mars orbiter: Falcon 9 or Falcon Heavy?

Compare Falcon 9 and Falcon Heavy for a Mars orbiter launching in the 2031 or 2033 window, and recommend a rocket, window and transfer. The spacecraft must reach a capture orbit about Mars with periapsis at least 250 km and apoapsis at most 33,000 km, verifying at least one full orbit, within 400 days of launch. It has a dry mass of 1,000 kg and a 2,000 kg propellant tank, of which 50 kg must remain as a reserve; its main engine gives 2,000 N at a specific impulse of 320 s. The rocket's upper stage performs the Earth departure burn; use NASA's published performance for Falcon 9 (drone-ship recovery) and Falcon Heavy (booster recovery). Show the options in each window, fully design and verify the key ones, and say what a faster transfer costs in delivered mass.

## Recommendation

Fly Falcon Heavy (booster recovery) in the 2031 window on the Type II transfer, launching 2030-12-14 and reaching the Mars capture orbit on 2031-09-26 (about 286 days). It was verified at 1,912 kg delivered, and `check_plan` passed on every phase and every requirement. Falcon Heavy can lift the full 3,000 kg spacecraft (capability is 4,797 kg at C3 12.5), so it is limited only by the capture burn. That makes the lowest arrival v∞ (3.44 km/s) the winner. Falcon 9 can lift only 2,220–2,400 kg at the C3 values it can fly, which forces you to leave 600–780 kg of propellant off the tank. Falcon 9's best verified result is 1,340 kg, so Falcon Heavy delivers 572 kg (43%) more in the same window. The 2031 window beats 2033 by 159 kg (9%) on Falcon Heavy and 90 kg (7%) on Falcon 9. If you need the shortest cruise, the Type I option (below) costs 130 kg.

## Verified designs

| Design | Window | Rocket | Type | Launch | Orbit insertion | Days | C3 (km²/s²) | Rocket capability (kg) | Separated (kg) | Propellant used (kg) | Delivered (kg) | Seed estimate (kg) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| mars-2031-falcon-9-asds-typeii | 2031 | Falcon 9 | II | 2031-01-09 | 2031-10-25 | 290 | 9.62 | 2,260 | 2,220 | 880 | 1,340 | 1,391 |
| mars-2031-falcon-heavy-recovery-typei | 2031 | Falcon Heavy (Recovery) | I | 2031-03-02 | 2031-09-26 | 208 | 18.27 | 4,059 | 3,000 | 1,218 | 1,782 | 1,846 |
| mars-2031-falcon-heavy-recovery-typeii | 2031 | Falcon Heavy (Recovery) | II | 2030-12-14 | 2031-09-25 | 285 | 12.51 | 4,797 | 3,000 | 1,088 | 1,912 | 1,966 |
| mars-2033-falcon-9-asds-typeii | 2033 | Falcon 9 | II | 2033-04-29 | 2034-01-27 | 274 | 7.62 | 2,470 | 2,400 | 1,150 | 1,250 | 1,339 |
| mars-2033-falcon-heavy-recovery-typeii | 2033 | Falcon Heavy (Recovery) | II | 2033-01-28 | 2033-10-17 | 263 | 17.49 | 4,157 | 3,000 | 1,247 | 1,753 | 1,826 |

Every design passed the independent SPICE/DOP853 check and the user's requirements; its ephemeris is in `oem/` (CCSDS OEM, ICRF, UTC).

## Seed maps

| Window | Rocket | Most mass (kg) | Launch | Days | Fastest (days) | Its mass (kg) |
|---|---|---|---|---|---|---|
| 2031 | Falcon 9 | 1,408 | 2031-01-04 | 286 | 286 | 1,408 |
| 2031 | Falcon Heavy (Recovery) | 1,966 | 2030-12-15 | 286 | 146 | 1,130 |
| 2033 | Falcon 9 | 1,336 | 2033-04-29 | 272 | 257 | 1,216 |
| 2033 | Falcon Heavy (Recovery) | 1,826 | 2033-01-29 | 262 | 187 | 1,211 |

Seeds: 5-day departure and arrival grids, prograde Lambert transfers, the rocket's curve capping the launched mass, an ideal capture into the orbit.

## Launch vehicle data

NASA Launch Services Program (LSP), Launch Vehicle Performance Website, hosted by the LSP Flight Dynamics Branch at Kennedy Space Center: https://elvperf.ksc.nasa.gov/Pages/Query.aspx (Performance Query, High Energy orbit class, C3 vs. mass), one query per C3 row; queried 2026-11-02 (site updated 10 February 2026). NASA Launch Services (NLS II) contract vehicle performance: separated payload mass assuming a standard payload adapter and separation system. The site cautions that NLS performance often differs from what providers advertise and that mission-unique requirements often need additional analysis.

- **Falcon 9 (Full Thrust, ASDS)** (CCSFS SLC-40; launch-asymptote declination within ±28.5 deg): C3 0: 3,310 kg, C3 5: 2,745 kg, C3 10: 2,220 kg.
- **Falcon Heavy (Recovery)** (KSC LC-39A; launch-asymptote declination within ±28.5 deg): C3 0: 6,690 kg, C3 5: 5,855 kg, C3 10: 5,130 kg, C3 15: 4,465 kg, C3 20: 3,845 kg, C3 25: 3,270 kg, C3 30: 2,740 kg, C3 35: 2,255 kg, C3 40: 1,805 kg, C3 45: 1,390 kg, C3 50: 1,005 kg, C3 55: 650 kg, C3 60: 320 kg.

## Not modeled

- Point-mass n-body dynamics on JPL DE440 (no J2, drag or solar radiation pressure); impulsive and constant-thrust burns on one engine.
- The launch vehicle is NASA's published C3 curve and launch-asymptote declination range: no ascent, launch-window or debris-compliance modeling.
- Capture into the orbit by propulsion only (no aerobraking); no navigation errors, dispersions or flight-readiness claim.
- Seed maps assume an ideal impulsive capture; verified designs fly finite burns, which lose a few percent to gravity losses.

## Claude's report

Recommendation: Fly Falcon Heavy (booster recovery) in the 2031 window on the Type II transfer, launching 2030-12-14 and reaching the Mars capture orbit on 2031-09-26 (about 286 days). It was verified at 1,912 kg delivered, and `check_plan` passed on every phase and every requirement. Falcon Heavy can lift the full 3,000 kg spacecraft (capability is 4,797 kg at C3 12.5), so it is limited only by the capture burn. That makes the lowest arrival v∞ (3.44 km/s) the winner. Falcon 9 can lift only 2,220–2,400 kg at the C3 values it can fly, which forces you to leave 600–780 kg of propellant off the tank. Falcon 9's best verified result is 1,340 kg, so Falcon Heavy delivers 572 kg (43%) more in the same window. The 2031 window beats 2033 by 159 kg (9%) on Falcon Heavy and 90 kg (7%) on Falcon 9. If you need the shortest cruise, the Type I option (below) costs 130 kg.

**Trade table** (all from `mission_status`; "Verified" means `check_plan` passed, and every row's requirements check also passed)

| Mission | Departure → capture | Flight days (cruise to capture / total plan) | C3 km²/s² | Capability / separated kg | Prop used kg | Delivered kg | Margins | Status |
|---|---|---|---|---|---|---|---|---|
| mars-2031-falcon-heavy-recovery-typeii | 2030-12-14 → 2031-09-26 | 286 / 287.4 | 12.51 | 4,797 / 3,000 | 1,088 | **1,912** | peri +39 km, apo +991 km, reserve +862 kg, launch mass +1,797 kg | Verified |
| mars-2031-falcon-heavy-recovery-typei | 2031-03-02 → 2031-09-27 | 209 / 210.6 | 18.27 | 4,059 / 3,000 | 1,218 | 1,782 | peri +31, apo +974, reserve +732, launch mass +1,059, declination 4.1° from the ±28.5° limit | Verified |
| mars-2031-falcon-9-asds-typeii | 2031-01-09 → 2031-10-26 | 291 / 291.8 | 9.62 | 2,260 / 2,220 | 880 | 1,340 | peri +31, apo +1,000, reserve +290, **launch mass +40 kg, C3 +0.38 under the table's limit of 10** | Verified |
| mars-2033-falcon-heavy-recovery-typeii | 2033-01-28 → 2033-10-18 | 264 / 265.0 | 17.49 | 4,157 / 3,000 | 1,247 | 1,753 | peri +38, apo +950, reserve +703, launch mass +1,157 | Verified |
| mars-2033-falcon-9-asds-typeii | 2033-04-29 → 2034-01-28 | 274 / 275.8 | 7.62 | 2,470 / 2,400 | 1,150 | 1,250 | peri +62, apo +906, reserve +200, launch mass +70 | Verified |

All five coast 1.32–1.33 orbits in the capture ellipse (peri 280–312 km, apo about 32,000 km). Total durations are 210–292 days, well inside the 400-day limit.

**Seed estimates (mapped with `porkchop`, not designed or checked; ideal impulsive capture)**

| Option | Dates | C3 | Seed delivered kg |
|---|---|---|---|
| FH 2031 Type II (best cell) | 2030-12-14 → 2031-09-27 | 12.4 | 1,966 |
| FH 2031 Type I, 197 d | 2031-03-08 → 2031-09-21 | 22.1 | 1,824 |
| FH 2031 Type I, 173 d | 2031-03-11 → 2031-08-31 | 26.3 | 1,651 |
| FH 2031 Type I, 158 d | 2031-03-11 → 2031-08-16 | 27.7 | 1,405 (launch-limited at 2,981 kg) |
| FH 2033 Type II (best cell) | 2033-01-28 → 2033-10-18 | 17.4 | 1,826 |
| FH 2033 Type I, 239 d | 2033-05-25 → 2034-01-19 | 12.8 | 1,664 |
| FH 2033 Type I, 202 d | 2033-06-08 → 2033-12-27 | 26.4 | 1,622 (declination −28.4°, at the limit) |
| F9 2033 Type I, 254 d | 2033-05-17 → 2034-01-26 | 9.8 | 1,246 |
| F9 2031 | none faster than about 289 d | – | The pareto front has only the 287–289 d Type II cells. |

**What a faster transfer costs in delivered mass**
- **Falcon Heavy, 2031:** 286 d → 209 d (77 days faster) costs 130 kg verified (1,912 → 1,782, about 1.7 kg/day). It also needs C3 18.3 instead of 12.5 and a −24° declination.
- **Going further:** seeds show 173 d (36 more days) loses about another 170 kg, and 158 d about another 250 kg.
- **No arrival-date gain:** the Type I launches 78 days later and arrives within a day of the Type II. It shortens cruise time, not the date you reach Mars.
- **2033 is less efficient:** the seed for 239 d costs 162 kg for 24 days, about 6.7 kg/day. In 2033 the fast cells hit the declination limit.
- **Falcon 9:** has no faster option inside its published C3 table.
- **Seed vs verified:** finite-burn capture loses 2.7–6.6% against the seed estimates, partly because of gravity losses and partly because Falcon 9's loads sit just below its curve.

**Things to know about the designs**
- **Falcon 9 margins:** the 2031 case has only 40 kg of launch-mass margin and 0.38 km²/s² of C3 margin to the table's limit of 10, so any slip makes it infeasible. The 2033 case has 70 kg of mass margin.
- **Falcon 9 2033 launcher burn:** it needed a −19.7 m/s out-of-plane component in the launcher burn, much more than the other cases, to put the n-body encounter in the right place. The solver also stalled twice before I restarted it from a hand-extrapolated point.
- **Periapsis aim:** I aimed the pre-burn periapsis high, because the finite capture burn pulls the post-burn periapsis down by 220–350 km.

**What is not modeled**
- Ascent, launch-site timing and azimuth, and the parking-orbit coast. Each departure is a launcher burn from a 200 km circular state, graded against NASA's published C3 / mass / declination table.
- Upper-stage burn losses and finite upper-stage burns.
- Navigation or injection errors, dispersions, and mid-course correction propellant. The 50 kg reserve is only a ledger check.
- Perturbations beyond point-mass gravity (no J2, drag or radiation pressure), and long-term stability of the capture orbit (only about 1.3 orbits verified).
- Aerobraking, aerocapture and orbit-orientation requirements (none were imposed).
- Fuel for a capture-orbit trim or a later periapsis raise.
- Flyby routes and multi-revolution transfers (only direct prograde Lambert cells), and the Falcon Heavy expendable variant (not requested).
- The seed maps use 2–3 day steps, so a better cell could exist between grid points.
- Flight readiness.
