sevenperforce

There's an idea. What's the value per gallon of water in orbit? And hey, that's something that would be enabling even without LOP-G. Go grab about a billion in venture capital, buy a reusable electric-pump-cycle hydrolox engine with accompanying thrusters, and build a solar-electrolysis propellant depot to be lofted into LEO by Falcon 9, then buy Falcon Heavy launches to send as much water as you can to the station, as frequently as it will launch. Use your hydrolox engine and thrusters to build a tug with a metallic heat shield for aerobraking, then offer direct-to-GEO services from basically any orbit. Masten has done a lot with small methalox thrusters that would be usable for a lander engine under the right circumstances.

The usable volume in a Falcon 9 fairing over 216 cubic meters. If you poured water into the fairing's usable volume, you'd only be 1/3 of the way full before you'd hit Falcon Heavy expendable's 63.8 tonne limit. For reusable configurations you'd use even less of the volume before you'd be mass limited. Now, if you're trying to loft hydrolox in bulk, then yes, you're going to run into problems filling up Falcon Heavy expendable's mass budget. A fairing-sized hydrolox tank would mass just about exactly 62 tonnes. But, again, we aren't flying Falcon Heavy expendable, and we aren't lifting pure hydrolox. Common bulk densities for reference: Water: 1 g/mL Hydrolox: 0.32 g/mL Methalox: 0.83 g/mL Kerolox: 1.03 g/mL Hypergolics: 1.2 g/mL

Looking at the Delta IV user guide, ULA offers about a dozen different Delta IV PAFs with suggested maximum load limits ranging between 5.0 for the flimsiest ones and 9.1 tonnes for the heaviest ones. The Atlas V has eight different PAFs with maximum loads ranging from 6.3 tonnes to 9 tonnes. The New Glenn user guide released two months ago and cited explicitly that they could do at least 20 tonnes in a dual-manifested launch, which makes that the lower bound of structural limitations on the upper stage (though I do not know if their PAF can handle 20 tonnes in a single-payload mission). The 45-tonne reference orbit was, naturally, inclusive of residuals. In terms of pure upper-stage structural limits, we know that Delta IV Heavy took the 21-tonne Orion EFT to LEO. So its stage structural limits are more than 230% of its biggest PAF.

Refueling isn't a magic fix for issues like boiloff and ullage. Starship gets around it by having such a massive square-cube advantage that boiloff and generous RCS use are inconsequential, but something like Centaur can't manage nearly so easily. IIRC, Centaur has to actively vent hydrogen boil-off throughout its entire coast just to maintain ullage and prevent the tanks from popping. The Falcon 9 user's guide suggests the current payload adapter is limited to 11 tonnes for center of mass reasons, but says that higher-mass payloads can be accommodated as a special service. I believe the problem is not so much with structural loads on the upper stage (and certainly not for the lower stage) as it is the bending moment on the payload adapter during pitchover under power. If the payload's center of mass is too high above the payload adapter then it will not be able to stay secure. The largest payloads launched in a fairing to date have been the Iridium NEXT clusters at over nine tonnes. But all you need to remedy that is to design a custom PAF, which isn't so bad. Structural limitations on the vehicle itself are probably much higher.

How far along are the XEUS plans for handling boiloff? That's always a challenge, but particularly with hydrolox. And a dual-thrust-axis lander makes the addition of a crew capsule rather challenging, particularly if the crew capsule is sized for independent abort. That's why I suggested putting a refuel module up on a separate LEO launch in order to rendezvous with the lander. Its propellants would be used both to refuel the ascent stage and to act as a drop tank for the descent stage, such that the descent stage can fill the role of tug. Plus, it will be light, meaning that there will be plenty of residuals in the launch vehicle upper stage to push both it and the lander onto TLI. Does anyone know the upper limit of discrete payloads for LEO by most commercial vehicles? Because if we know that number then we can start sizing accordingly.

Logistically speaking... There is value in having hypergolics on the ascent stage, so let's assume an ascent stage powered by either a single SuperDraco or clusters of smaller hypergols. It will need to dock to the lander stage with capacity for propellant transfer. The smaller the better, for pretty much every reason. This also means the lander will need to carry hypergols for refueling the ascent vehicle, though not necessarily for itself. For your mass ratio, you want to shed spent mass whenever possible, and for complexity/reliability you want to drop stages as infrequently as possible. The ideal scenario is to have your ascent module refuel tanks bolted to the same jettisonable module as your lander's lunar orbit insertion tanks (since it's braking itself into cislunar space) so you only have one separation event. For dV reasons, let's assume cryos for the lander. Scenario: Crew and unfueled ascent module are already at LOP-G. Commercial launch 1 places lander module in LEO with just enough propellant for lunar deorbit and landing. Commercial launch 2 takes refueling module into LEO but does not separate. Lander executes rendezvous and docking with refueling module, then Commercial launch 2 restarts its engine for TLI burn. Lander completes TLI burn (if necessary), coasts, and then brakes into cislunar orbit at LOP-G, drawing from cryo tanks in the refueling module. Crew enter ascent module and dock to lander. Ascent module is refueled from refueling module. Refueling module is jettisoned either at the LOP-G or once in LLO. Descent stage lands and the surface mission is conducted. Ascent module returns to LOP-G to repeat the cycle. This expends only one BLEO engine assembly and achieves a full lunar landing in only two commercial launches rather than three. Because the lander is sent to LEO, it can be dramatically larger than otherwise possible. With Atlas V 551 it could go up to 19 tonnes; with Delta IV Heavy, 28 tonnes; with Falcon Heavy reusable it could go up to 30 tonnes or more. New Glenn, Vulcan, or Delta IV Heavy would be ideal for launching the refueling module because the residual hydrolox will give better TLI performance.

Merlin Vacuum without nozzle extension: Merlin Vacuum Nozzle Extension by itself: Combined: It's the same deal with the Kestrel. For the Kestrel, here you can see both the ablatively-cooled chamber and throat attached to the radiatively-cooled nozzle:

It's a Kestrel without the nozzle extension attached.

They need to either dust off the Kestrel and convert to methane-GOX or do that with the SuperDraco.

I don't think I've ever seen a that picture of the Kestrel before.

Spent entirely too long digging into the nine companies tapped by NASA for CLPS to expand the roundup of currently-available engines. Here's what I came up with. BE-3U (Blue Origin), 530 kN. Hydrolox. Broadsword (Masten Space Systems), 160 kN. Methalox. Reaver 1 (Firefly Aerospace), 182.2 kN. Kerolox. RL-10C (Aerojet Rocketdyne), 110 kN. Hydrolox. SuperDraco (SpaceX), 90 kN. Hypergolic. Lightning 1 (Firefly Aerospace), 70.1 kN. Kerolox. AJ-10 (Aerojet Rocketdyne): 27 kN. Hypergolic. HD5 (Intuitive Machines), 24 kN. Methalox. Rutherford (Rocket Lab), 24 kN. Kerolox. Katana (Masten Space Systems), 18 kN. IPA+LOX. PECO (Moon Express), 5.64 kN. RP1+HTP. Machete (Masten Space Systems), 4.4 kN. IPA+HTP. FME (Team Indus for ORBITBeyond), 0.44 kN. Hypergolic. Deep Space Engine (Frontier Aerospace for Astrobotic), 0.44 kN. Hypergolic. Draco (SpaceX), 0.4 kN. Hypergolic. A slightly broader spread, but still not a lot to choose from. It is a shame that SpaceX's gas-gas methane-oxygen thruster has been put on the back burner. Imagine a SuperDraco with a regenerative cooling loop that pre-vaporizes both LOX and CH4, feeds a small amount back to keep the tanks pressed, and then feeds the rest into the chamber with a simple spark igniter. Gas-gas engines are simple, reliable, and while it wouldn't have the best TWR in history, it would have superb efficiency and relatively decent propellant bulk density. NASA's plan currently requires that all vehicle components be limited to what can be sent to TLI by commercial launch providers AND fit inside a five-meter fairing. It's really rather restrictive. Thinking back to Constellation...I would love to see a single-stage transfer, refuel, and descent/surface module which is assembled in LEO via two commercial launches prior to TLI. As it is, the NASA approach requires no less than three commercial launches, plus crew via SLS, for every moon landing, and that's without any supplies being landed on the moon. At least then you would only have a single disposable powerplant braking into cislunar space for each mission instead of literally three.

A human-sized lunar lander needs to start with an appropriately-sized engine...so, what's out there? I thought I would do a quick roundup of engines that could conceivably be used or adapted for use in cislunar space.... BE-3U: 530 kN RL-10: 110 kN SuperDraco: 68-90 kN Rutherford Vac: 24 kN* AJ-10: 27 kN Not a big list, and not much to work with. Rutherford is included because although kerolox is problematic for ignition reasons, the cycle is fantastic for cislunar ops. How many did I miss?

We know the Falcon numbers to a high degree of accuracy. We'd need to estimate for NG based on published imagery and propellant bulk density. Bezos is tight-lipped. To the OP: there is an optimization function for any proposed mission profile or configuration set, but there are a lot of variables. A vehicle intended primarily for launching BLEO will ten to be underpowered for large LEO payloads, and vice versa. Plus, if you use different propellants on the first and second stages, you get an entirely new set of variables. Recovery also complicates.

The two upper stage engines will be BE-3Us at 530 kN each, so 15.8:1.

Second stage uses a Merlin 1D Vacuum at 934 kN; SL thrust on the first stage Merlin 1Ds is 845 kN.

Once they get it working reliably with one, they can add another ship to catch the other.

As @Rakaydos said, they want as much drag as possible on entry. The grid fins on Falcon 9 are used because they need fine guidance and short-chord control surfaces perform better without stalling at supersonic speeds. Traditional fins (a la New Glenn) take more torque to move and stall more easily while supersonic. In contrast, the "flaps" on Starship will be operating initially at hypersonic speeds, where everything is in a state of stall and you are using differential compressive drag to induce roll and pitch rather than any fine aerodynamic guidance. They will function less like canards and more like airbrakes.

I do like some of the ideas dropped by posters about more dynamic moons. Honestly, adding more interesting moons would be the best reason to add another planet. More surfaces mean more stuff to do. I'd like to see a Saturn/Neptune analogue between Eeloo and Jool with at least two moons. A very close, inclined, eccentric moon without an atmosphere but with active volcanism due to tidal flexing. Volcanism would be represented by solid regions of glowing "lava" in the ground snaking down mountainsides and "pooling" in flat lakes...the "lava" would emit enough light to power a solar array but would be too hot to stand on. It would also be cool to implement something like an earthquake; landers would need to have rover wheels to deal with the shock. It would be good to see some challenging terrain including cliffs, overhangs, etc., and update the Klaw so it could be used to cling to the side of a cliff and construct bases that way. Outside the orbit of the volcanic moon you could have "rings" consisting of dozens of very small asteroids, procedurally generated. Making the ring area a "space biome" and allowing a way to refuel xenon tanks in that biome would also be amazing...you could make the ring biome darkened so that no solar array would function, just to make it an added challenge. Farther out, an icy moon with a very whispy oxygen atmosphere, cryovolcanism at the poles (a surface that is semitransparent and has 100% ore concentration but offers reduced traction) and fissures with water lakes near the equator. Underwater caves would be insanely cool. Of course these would be orbiting something that is basically just "Jool Lite" which isn't much fun. If they really wanted to make it interesting, they could give it an atmosphere even soupier than Eve, but only a few kilometers thick, nearly-zero visibility at sea level, with a shallow ocean underneath. Instant death if you try to descend very far into the ocean, but it's something you could at least land in/on. Major challenge but achievable. Could give the atmosphere a small amount of oxygen, or not if desired.

Seems doubtful to a first order.

There are tethers; they're called ladders. Accidentally let go and you're dead, unless your ship can maneuver to you.

Nothing so fancy; it can be done in a much more intuitive (albeit crude) fashion. Just use segmented boosters. Add two new parts: a cylindrical "SRB segment" and a stubby "Segmented SRB nozzle". You can stack segments on top of each other and each one functions like an independent SRB in the VAB (with thrust limiter and fuel load sliders, etc.), but when fired, their thrust originates at the attached nozzle. You could even have multiple nozzles based on utilization (TVC, vacuum, fixed SL). If you want to preprogram your thrust to vary, then you merely need to set some of the segments to burn their fuel faster than others. Thus they will burn out in sequence, which steps down your total thrust over the course of the burn. Then the entire assembly is jettisoned at once.

Damage is often a function of shear force/pressure rather than gross kinetic energy or momentum.

Very Kerbal indeed, but I love it. The required amount of Ga to pull it off could be prohibitive. Why do you say that mechanically grinding a tank on the moon would be impossible? A robot with a laser cutter could pretty easily slice an aluminum tank into narrow ribbons which could then be fed into an paper-shredder-sized grinder. Enrichment of the iron oxide or other soils could be difficult. Regular proton bombardment has resulted in highly reduced soil, with most iron on the moon either being metallic or +2 oxidized instead of +3 oxidized as is standard on Earth. Regular iron oxide is not ferromagnetic, either, so a mechanism for enriching the soil (if its oxygen weight was insufficient to support combustion) would be challenging. That, I think, would be the limiting variable.

They really did.

Mockup of what this architecture might look like after several sequential missions but before any octagrabber-related base-connection activity: Note that each descent module uses a hab module as its core with tanks and engines on the sides, a docking port in the front to mate with a pressed rover, and an airlock in the back. Contingency abort vehicle visible over on the right.