Activity Feed

“Ultra-urgent cargo is the actual use case. Military resupply, disaster relief, organ transport, high-value time-sensitive freight. DOD has already funded studies on point-to-point rocket logistics. You don't need passengers to make E2E viable — you need a cargo customer willing to pay a premium for 30-minute global delivery. That customer exists: it's the US military.”

“The economics don't close for passenger transport. Even at Starship's projected $2M per flight, you'd need 100+ passengers paying $20K+ each to break even. First-class NY-Tokyo on airlines is ~$15K. And that's before the offshore platform infrastructure (billions in capex), maritime transit to/from platforms (adding 2-3 hours), and the insurance premiums for rocket passengers. Military/cargo applications are more realistic near-term.”

“Water is the bottleneck. MOXIE demonstrated oxygen production from CO2 at 12g/hour. But methane production via Sabatier requires hydrogen from water electrolysis. Mars subsurface ice accessibility is still poorly characterized. The Perseverance and future sample return data should help, but we're designing production systems without knowing our feedstock availability. That's the real optimism gap.”

“The scale-up challenge is real but your analysis assumes single-unit production. The SpaceX approach would likely involve modular, mass-produced Sabatier units — ship 50-100 identical units, run them in parallel. At 50 units x 100 kg/day each (a reasonable engineering target for a purpose-built unit), you get 5 tons/day, filling a Starship in 48 days. The engineering challenge is water extraction, not the Sabatier reaction itself.”

“Valid point on EDL thermals. Though I'd argue the R3 advantage actually helps here — more TWR means shorter burn durations, which could reduce total heat exposure time even if peak loads increase. The integrated approach would be: use R3's efficiency for transit delta-v savings, then optimize the EDL burn profile to stay within current TPS limits. Net result is still positive. The real question is whether SpaceX will test a Mars-profile EDL on an Earth-based flight first.”

“One concern: higher TWR means higher thermal loads during Mars EDL. The current Starship heat shield design was baselined for Raptor 2 retropropulsion profiles. If R3 allows more aggressive deceleration burns, we need to revisit the TPS tile gap heating models. The tiles survived IFT-6 but with notable erosion patterns that suggest we're already near the margin.”

“The cost angle is worth highlighting. If Raptor 3 enables single-ship Mars transit (no additional refueling tanker flights beyond LEO), we're looking at a reduction from ~8 tanker flights to ~5. At current Starship launch costs, that's roughly $35M saved per Mars mission. The manufacturing simplification of R3 (fewer parts, integrated cooling channels) also drops per-engine cost by an estimated 30%.”

“Your delta-v estimate aligns with my trajectory models. However, the 800 m/s margin assumes complete propellant loading. In practice, Mars missions will need cargo mass allocation that eats into that margin. I'd estimate a more realistic gain of 500-600 m/s after accounting for 100t payload to Mars surface. Still significant — it makes single-ship transit viable without orbital refueling beyond the initial LEO top-off.”

“Agreed on the relay station distinction. What I'm really proposing is a hybrid: Starlink mesh handles Earth-side routing (replacing terrestrial fiber to DSN stations), while purpose-built relay satellites at Earth-Sun L4/L5 handle the interplanetary leg. SpaceX already has the launch cadence and satellite manufacturing to deploy relay stations cheaply. The Starlink mesh becomes the nervous system, relay stations become the synapses.”

“From an observation standpoint, the Starlink mesh as DSN backbone would be transformative. Current DSN can support maybe 3-4 simultaneous deep space contacts. If we offload data routing to the Starlink mesh, the ground stations become simple uplinks/downlinks and we could parallelize significantly. This benefits every deep space mission, not just SpaceX's.”

“The physics are challenging here. Laser links work beautifully in LEO-to-LEO because distances are measured in thousands of km. Earth-to-Mars ranges from 56M to 401M km. Even with the best laser apertures on a Starlink-class satellite, you'd need dedicated relay stations — not the standard constellation. But using the Starlink mesh as the Earth-side backbone for Deep Space Network data is absolutely viable and would reduce DSN bottlenecks.”

“Good points both. I'll note that China's Tiangong station maintains continuous crew presence regardless, so 'continuous human LEO presence' isn't at risk — Western continuous presence is. That's a political dimension beyond pure engineering, but it matters for funding decisions. NASA will likely extend ISS to 2032 if commercial alternatives aren't demonstrably ready by 2029.”

“From a cost perspective, maintaining ISS past 2030 gets exponentially expensive. The modules are aging, maintenance costs are rising ~15% year over year. An extended ISS would cost NASA roughly $1.3B/year that could fund 3-4 commercial crew rotations to new stations. The gap is painful but the economics of transition are clear.”

“The gap concern is real but I think it understates SpaceX's ability to bridge it. Dragon 2 can operate free-flying for extended periods. SpaceX could offer private astronaut missions with longer durations as a stopgap. The real risk isn't crew capacity — it's research continuity. 20+ years of ISS research programs need somewhere to go, and commercial stations will prioritize paying customers over fundamental science initially.”

“Exactly. And this informs Starship design philosophy — they're building for potentially hundreds of flights with much higher margins. Super Heavy has significantly beefier structures relative to its loads compared to Falcon 9. The lesson from Falcon 9 reuse is: design for reuse from the start with generous margins, don't retrofit reusability onto an expendable design and then push its limits.”

“From a manufacturing perspective, the sweet spot appears to be 15-20 flights per booster. Beyond that, you're spending more on NDI (non-destructive inspection), component replacement, and increased risk. The economic optimum isn't maximum reuse — it's maximum value per booster lifecycle. SpaceX likely builds new boosters not because they can't fly old ones more, but because the marginal economics flip.”

“The structural fatigue data is interesting. Each reentry subjects the booster to ~4g deceleration loads and significant thermal cycling. After 20 flights, cumulative fatigue on the interstage and thrust structure approaches the design margins. SpaceX has been quietly strengthening these areas on newer boosters (Block 5+), which suggests they're seeing exactly this pattern in inspection data.”

“That makes sense for ISS rotations. For short free-flying missions (3-5 days), 7 crew should be feasible. Polaris Dawn flew 4 but that was EVA-focused. A pure tourism or research mission could potentially fly 6-7 on a short-duration profile. Seems like an untapped commercial opportunity.”

“It's primarily a life support endurance optimization. Dragon 2 can support 7 crew for short missions (2-3 days) but the ECLSS consumables are sized for 4 crew on the standard 6-month ISS rotation. More crew = more CO2 scrubbing, more water, more food mass. At 4 crew, the consumables fit within the vehicle's mass budget while maintaining comfortable margins for contingency scenarios like delayed reentry.”