sevenperforce

It seems they've done a combo hard-suit/soft-suit with that stiff coupled joint at the waist. Smart. But man, I still feel like there's got to be a way for form and function to interlock a little better.

Is that a suitport with an extra attachment, or no? Looks painful. Elon to tweet "sexy EVA suit" concept in 3, 2, 1...

ISS has been extended to 2030 so it's possible, though I think they want to get away from ISS dependency. I was re-reading the HLS Concept of Operations publication and noted that they allow for up to 120-day transits to NRHO for uncrewed modules to save dV. An ion crawl from the ISS to LOP-G would cost about 7.2 km/s, which is not unprecedented -- Deep Space 1 achieved over 7 km/s and Dawn achieved over 11 km/s. However, that velocity change by Dawn took nearly six years of thrusting. A fully-stacked lunar sortie vehicle would be more than fifty times the mass of Dawn. If we want the lander stack to make the journey from LEO to LOP-G in under a year, we would need 630+ ion thrusters and a 2.1 MW solar array. For reference, that's more than twenty times the power generation capability of the main ISS solar array. The HLS ConOp did mention that there would not be a Canadarm at LOP-G during the initial phase, which complicates the use of expendable drop tanks.

Foetuses do not drop solely due to gravity, but their weight is critical to stimulating contractions. nethersl birth would require induction. But of course no one would ever, ever countenance the slightest chance of pregnancy (at any stage) on orbit. Obligatory:

Wow. This suggests a 460+ kg nozzle extension. Is that normal? Wow again. that leaves a lot on the table. Anyone know the lowest pressure you can hit at SL without flow sep?

Exactly. If there is any mission-critical component you want to replace each time, it's the ascent engine. The more I think about it, the more convinced I am that replaceable, expendable, pressurized hypergolic drop tanks present the lowest barrier to realizing near-term component reuse. Someone, probably LockMart or ADS, would need to develop a reusable structural coupling that would be able to transfer propellant at high flow rates and handle significant force loading. Then they just need a lightweight pressurized hypergolic+helium tank and a grapple fixture on the back end. A prototype (subscale tank, full-scale coupling) could easily be sent up to the ISS for Canadarm testing. Such a tank could be used to extend Orion's capabilities (e.g., letting it act as a lifeboat to rescue a stranded ascent module in LLO), re-equip a "space taxi" ascent module, and so forth. Tankage weight is the least reusable part of the Artemis architecture, because a refill supply ship HAS to carry the props somewhere. The only disadvantages -- you can't regeneratively heat the helium pressurant, and you don't have quite as much of a square-cube advantage as other systems. Otherwise it is the most sustainable, mass-efficient way to progress cislunar reuse until we have cryogenic ISRU. With such capabilities, you could theoretically use distributed launch here in LEO to do each lunar lander mission with just a single commercial TLI launch. Send up a "pallet" of the expendable tanks on an Atlas V or a Vulcan, then send up a new lander/ascent module on Falcon Heavy to rendezvous while still connected to the Falcon upper stage. Autonomously mate, push as far toward TLI as possible, and then let the lander/ascent module do the rest of the TLI (plus LOP-G rendezvous) using props from the expendable tanks.

The one place it might make vague sense is refilling the ascent vehicle. The tanker does not need a proper engine, after all (it can rendezvous with LOP-G on RCS alone), so you sorta save the weight of a single engine, as opposed to replacing the tanks and engine each flight (that's if you plan on reusing the ascent vehicle anyway). However, that's also the MOST critical engine of the whole affair, so you don't really want to reuse it too many times. It's not too bad. There are zero-cost trajectories away from NRHO that impact the moon or enter heliocentric orbit; they just take a long time.

Ditto on refuel/refill. I suppose I could say "repropellantize" if I wanted to be annoying. Most of what you need for repeated lunar mission activity is engines and propellant. So shipping new engines and propellant just to refill existing engines and propellant is a waste. Reusing the transfer vehicle makes no sense whatsoever. Reusing an ascent vehicle doesn't make much sense because you can have a reusable cabin and simply mate a new booster module to it each time. The only way it makes sense is if you are reusing a dedicated lander and want to avoid shipping new ground structure/landing legs. But there are no plans to reuse the lander, and you wouldn't want to anyway because it represents the highest amount of unnecessary dry mass and the highest risk factor for damage (due to debris impingement, etc.) and it's where you want to attach new ground support supplies anyway. Debris impingement is the one flaw I foresee with my transfer-ascender-cabin plan.

To be fair, it is confusing. The older mission specs have the transfer vehicle returning independently to LOP-G to be refueled by a logistics launch, while the newer mission specs have it smash itself into the moon. But it certainly will not stay in LLO. As he points out in the video, it would need to do a hella plane change, which it won't have the dV for. Reusing the tug just doesn't make sense to me. Why send a refueling mission with tanks and an engine when you could just send a new tug?

Scott also makes the mistake of thinking the transfer vehicle is intended to stay in LLO as a tug following the mission. Replacing the "transit stage" with a fuel tank is a great idea; it just requires mature, large-scale propellant transfer tech, which is something we simply don't have developed well enough yet. If you have full propellant transfer then you have access to a lot of different architectures.

I feel like the presumed reliability of pressure-fed hypergolics will be a strong argument in their favor, at least for the ascent vehicle. But proven flight heritage can change that. If Bezos has delivered dozens of payloads to the lunar surface without a hitch on turbopumped hydrolox, the possibility of using the exact same platform as a landing-and-ascent module underneath a reusable crew capsule becomes much more attractive.

Thinking about the crew ascent vehicle also being an independent cargo descent vehicle reminds me of the Blue Moon lander. Braked into LLO by the residuals of a New Glenn upper stage, it can deliver up to 4.5 tonnes to the lunar surface. Accordingly, it is almost exactly the right size to act as a lander/ascent module for a crew capsule. Dual-expander hydrolox is not ideal in comparison to pressure-fed hypergolics but it would be much lighter.

You actually end up with better performance, not worse performance, if you eschew propellant transfer altogether for expendable, replaceable tanks for the crew capsule. Instead of carrying 1.1 tonnes of propellant in its tanks to transfer into the capsule, the lander module would carry a couple of pressurized tanks on its sides that would be detached via Canadarm and mated to structural umbilicals on the sides of the capsule. It only saves a little mass -- the weight of what would otherwise be empty tankage on the lander module -- but it's something. It also gives some additional options. For example, you could have a "stretched" version of the tank on hand with enough props to give the space taxi dV to LLO and back -- enough to rescue a stranded capsule. On the other hand, if you just love microgravity propellant transfer and want to go all in, you can make the monolithic crew vehicle/landing module/ascender fully reusable. Send it up on a Falcon 9 or an Atlas V 521 and let it go to the Gateway under its own power. Then all you have to do is find a way to get 9 tonnes of propellant to the Gateway (plus your transfer vehicle) whenever you want to do a lunar mission. One reason to prefer the reusable crew capsule approach, whether with onboard propellant or replaceable tanks, is that it offers a commonality element for unmanned lunar deliveries. Cut back to a single engine and load it up with cargo atop a Falcon Heavy, and it can drop serious tonnage on a one-way trip to anywhere on the lunar surface. Using the same lander for cargo descent and crew ascent would be tough to pull off under any other circumstances, but it really works here.

Using the more precise GR&As above, I'll try to nail down the numbers for the above architecture a little more cleanly. Before, I said a 3-tonne crew capsule with 800 kg for crew, cargo, and consumables. Assuming that the lander/ascent vehicle comes off of TLI at exactly 15 tonnes (since that's our peak throw) and develops 316 seconds of isp during the powered lunar flyby and rendezvous with the Gateway (summing 428.5 m/s), it will arrive at Gateway having burned 1.9 tonnes of propellant. It will take on 800 kg of crew, cargo, and consumables; its engines will next fire for the terminal landing burn. I earlier estimated it would do about 350 m/s of the terminal burn. That may go up or down, but it's a good starting point, so at touchdown it should mass around 12.4 tonnes. By the above metric, we'll add about 1.1 tonnes for the landing legs and ground structure. We have better materials and construction nowadays so I feel like that bakes in some conservatism. Assuming we want at least a third of a gee for efficient takeoff, we will need at least 45 kN. We could go with a single vacuum-expanded Superdraco, but I don't think it would be able to throttle low enough for hover. Better to do a tight cluster of three OMS-heritage throttleable AJ-10 engines at around 26.7 kN each and 319 s; with a decent gimbal range that should give single-engine-out capability early in the flight and double-engine-out capability later in the flight. This tells us our engine mass will be 0.3 tonnes. Going back to our original 15-tonne vehicle coming off TLI, that's 10.6 tonnes of tankage and propellant. Using Transtage's 14.5% tankage mass fraction gives us 1.5 tonnes of tankage and 9.1 tonnes of propellant, dry (loaded) mass of 6.7 tonnes. Now we can start at the end and work backward. It burns just under 1.8 tonnes to get from LLO to Gateway and around 7 tonnes to get from the surface to LLO, leaving nothing for the terminal landing burn or the rendezvous with Gateway. So we are not that far off, but it doesn't quite work. However, there's one puzzle piece that I think should make everything work. All other things being equal, it really makes sense to use a reusable crew cabin/capsule. In fact, it may well be the precise puzzle piece that makes everything work. Consumables can be easily swapped out (bottled air, CO2 scrubbers, and so forth) via Orion trips. And it can't be expended en route; it must make it back to Gateway...so why not keep it at Gateway? Saves having to build and send a new one every trip. The Zvezda module has been refueled numerous times by Progress and Progress M1 resupply spacecraft and has remained on orbit for two decades. Accordingly, it is not a huge leap to develop the capacity to have the crew cabin/capsule with RCS thrusters (think Dracos) that can be refueled from a docked or berthed vehicle. The mass of the RCS thrusters is negligible, and the tankage mass ratio would probably be similar to the 17.3% of the AVUM upper stage on Vega. The crew capsule would be able to make the journey from LLO to the Gateway on its own, entirely on internal propellant. To get 730 m/s on likely lower-efficiency RCS thrusters (I will use 300 s, since that's what the Dracos develop), you burn just over 1.1 tonnes of propellant to get from LLO to Gateway and the dry loaded mass of the crew capsule is right at 4 tonnes. Note that this also means the crew capsule can deliver itself to the Gateway easily, as Falcon 9 can throw five tonnes to TLI without even expending the core. Of course, you have to get that 1.1 tonnes of propellant from somewhere. The easy solution is to get it from the expendable lander/ascent vehicle, which will berth or dock to the base of the crew capsule rather than docking directly to Gateway. Without those three tonnes of extra mass being thrown from TLI, its same 15 tonnes includes 2 tonnes of tankage and 11.6 tonnes of propellant. It arrives at Gateway with 9.7 tonnes of propellant and transfers 1.1 tonnes to the capsule, leaving it with 8.6 tonnes. It will still take it around 7 tonnes of propellant to get off the surface and into lunar orbit, but that leaves 1.6 tonnes of propellant for the terminal landing burn: around 300 m/s. So, what are the requirements on the transfer vehicle? Well, the transfer vehicle needs to deliver itself from TLI to the Gateway (428.5 m/s), and then it needs to deliver 17.2 tonnes from the Gateway to LLO (730 m/s), and then provide the first 1,570 m/s of the descent and landing burn. This is where the RL-10 can really shine. A 15-tonne hydrolox transfer stage, launched onto TLI by anything capable of throwing it, will burn 1.4 tonnes of propellant getting from TLI to the Gateway. Pump-fed hydrolox tanks boast a mass fraction of around 9%, meaning the stage will dock to the lunar sortie stack with a dry mass of 1,529 kg and 11,855 kg of propellant...giving it 2,165 m/s of dV. That's enough to get it from Gateway to LLO and from LLO to about 435 m/s short of the lunar surface...within 135 m/s of where we need to be. So it's doable. Two commercial launches, one reusable "space taxi" crew capsule.

The most critical question in evaluating possible architectures is fluidics management. On one extreme you have proposals like the LockMart sixty-tonne fully-reusable elevator monstrosity, which depends on a cislunar hydrolox depot supplied by a solar electrolysis plant, or Starship itself. If you have total propellant transfer capability, all kinds of unique architectures open up. Of course, that does not play well with pressure-fed solutions, so it excludes a lot of current/COTS planning and requires ZBO technology. The other extreme is to have no propellant transfer whatsoever. The only way to do reuse under these circumstances is to have a reusable crew cabin with an expendable bus, or at most a system that uses expendable, replaceable drop tanks which can be attached using a Canadarm to structural umbilicals. The middle ground is to have limited propellant transfer, slightly more flexible than what we have already demonstrated with Progress, etc. In this regime, it is possible to have a reusable vehicle with small pressure-fed tanks that are vented to vacuum and then filled via umbilical from much larger, expendable tanks. This has some advantages over a zero-transfer approach, though they are not necessarily overwhelming advantages. Since the larger tank has to be expended anyway, the launch supply requirements aren't dramatically different. Still, it can be a useful way to achieve certain solutions. For example, if you used a crasher stage for transfer and descent and an expendable landing/ascent stage but wanted to have a reusable crew cabin, you could give the crew vehicle ~800 m/s of onboard tankage, refilled each sortie from the landing/ascent stage. Then you can leave the spent ascent stage to decay in LLO and take the crew vehicle back to the Gateway on RCS. Another question is the number of permissible separation events. Being KSP players, we are highly inclined to shed any unnecessary mass with extreme prejudice. For example, in a combined lander/ascender architecture, we would be strongly inclined to jettison the ladder, landing legs, and any compressive structures at liftoff. Such a design then thoroughly trounces the Apollo descent+ascent solution, because not only do you shed unnecessary mass before liftoff, but the surface structures are far lighter since they have less mass to support. However, separation events add risk, which may not be acceptable in real life. Several ground rules and assumptions need to be established. The mass fraction of tankage should be estimated by comparison to a launch vehicle stage of commensurate size and engine cycle as well as identical propellant. An expander-cycle hydrolox stage will have a much poorer tankage mass fraction than a gas-generator kerolox stage, which will have a much poorer tankage mass fraction than a hypergolic turbopumped stage, which will have a much better mass fraction than a pressure-fed hypergolic stage. Pressure-fed stages are not quite able to get square-cube advantages, while turbopumped stages generally are. Estimating the additional mass of landing legs and associated ground structure is best done by reference to the excess mass of the Apollo descent module. The descent module had a dry mass of 1,983 kg and carried 8,200 kg of hypergolic propellants. 179 kg of its dry mass was, of course, its engine, but pressure-fed tanks for that much propellant should have only massed around 1,400 kg (compare to Transtage, with pressure-fed tanks massing 1,750 kg and 10,297 kg of hypergolic propellant). Thus we can estimate conservatively that the ground structural mass of the Apollo descent stage was no more than 590 kg. Since it supported a landed mass of 6,834 kg, we can add a 9.4% structural fraction to any lander.

I will try and put together a table. There are a few major considerations, including whether you have no propellant transfer, slight propellant transfer, or full propellant transfer. Also whether you have cryos, ZBO cryos, a combo of storables and cryos, or pure storables. I need to do some stage mass calculations to nail down a better idea of what sort of mass fraction is possible. My gut is that the best combination, all-around, will be a hypergolic ascent stage and a cryogenic transfer stage. But we will see. Yes, exactly. Basically you transfer extra dV to the ascent stage and thus end up with roughly equivalent-mass stages. The degree of reuse is a big deal as well. Unless you are able to do propellant transfer of some kind, reuse is of limited value. There are still a few options. For example, you can make the actual crew capsule reusable and expend the ascent stage and tanks in the vicinity of the Gateway after each flight. Saves a good bit of mass between sorties. Or, if you can do only limited propellant transfer, you can put RCS on the crew capsule and give it just enough tankage to get from LLO to Gateway, and expend the ascent stage in LLO.

Here's a novel thought. If fifteen tonnes is the maximum anything but SLS will be throwing to TLI, what about a fifteen-tonne landing module? Hear me out. Put the loaded hab at 3.8 tonnes: half the landed mass of Orion. Suppose crew, cargo, and consumables come to 0.8 tonnes, making the empty hab 3 tonnes. I'll use a mass ratio approximation from the Apollo descent module. Propellant mass was 8200 kg; dry mass 2134 kg. That's a structural fraction of 20.1% which can definitely be beaten by modern materials, shared components, and the square-cube law. Allowing ourselves 15 tonnes of total vehicle mass without crew, cargo, or consumables gives us a 12 tonne propulsion module; if we only use the square-cube law then that gives us 2358 kg of propulsion module dry mass and 9642 kg of props. At a notional isp of 320 s, the lander would have 2.95 km/s of dV. What's that good for? Well, it's good enough to provide the last 350 m/s of terminal descent, plus ascent, plus return to the gateway. It would need a transfer stage to ferry it, not just from the Gateway to LLO, but from the Gateway to LLO to the initiation of terminal descent: about 2.25 km/s. You need about 15 tons of hypergolics to push 15 tonnes of payload through 2.25 km/s. So you don't need three stages at all. All you really need is to beef up the transfer vehicle and let the ascent vehicle perform the terminal landing burn, and split your vehicles into two 15-tonne chunks. At that point, you can do two commercial launches and be done with it. You'll need a separate logistics launch to refuel both vehicles after they rendezvous with the Gateway, but those props are not super significant and you need the logistics launches anyway. If the ascender/lander can be reused, then you'd refuel it and the transfer vehicle from a dedicated logistics launch, which would mean two launches per sortie, ad infinitum. Alternately, you can stack both transfer and lander on SLS Block 1B Cargo and do it Constellation-style (albeit with rendezvous at Gateway rather than in Earth orbit). With that approach you have enough margin to not need a separate logistics launch.

Right, that's what I initially had issue with. Thought it could have been an error on NASA's part. Which would not have been improbable....

That's not what I read at the above article. However, when I look at the HLS Concept of Operations linked here, it suggests otherwise. Transfer vehicle is disposed by crashing into the moon; ascent module is disposed in the vicinity of Gateway. Initial surface ops are expected to have only two crew once again. Dry mass estimate for the ascent module still holds. It's interesting that in the three-stage configuration shown in the Concept of Operations, it explicitly has a Falcon Heavy silhouette launching all commercial payloads.

Very useful. If the transfer vehicle is supposed to ferry the ascent module back to the gateway, then there should be variation in the mass of the ascent module between the two-stage and three-stage options. Specifically, if a 9-12 tonne ascent module is using hypergolics with a notional Isp of 320 s, it would need 5.1-6.7 tonnes of propellant to get back to the gateway, putting its dry mass at 3.9-5.2 tonnes. Going only to LLO to meet the transfer vehicle is much cheaper and would cost only 4.2-5.4 tonnes of propellant, making the total wet mass only 7-10 tonnes. Establishing the expected dry mass of the ascent module at ~3.9-5.2 tonnes is interesting. Represents a ~100% mass growth over the Apollo ascent module and lunar hab, which is not THAT much given that it is supposed to have 100% more crew, and roughly half the weight of a full Orion plus tankage, as I indicated above.

Hence not using a runway landing.

LockMart has proposed the use of the Orion pressure vessel, ECLSS, and avionics for their notional single-stage lunar lander. https://www.lockheedmartin.com/content/dam/lockheed-martin/space/documents/ahead/LM-Crewed-Lunar-Lander-from-Gateway-IAC-2018-Rev1.pdf

I would hazard that the Starship has a considerably lower L/D ratio than the HL-10. Whether it would be enough to execute a landing flare is anyone's guess.

There's really a TON of stuff on there that's useless in cislunar space. I wonder how low it could go. Still doesn't solve the airlock problem.

I know the titanium heat shield carrier structure alone masses over 1300 kg on Orion. I cannot find mass estimates for the thermal backshell or the heat shield itself, anywhere. Nor for the parachutes. If we make the gross estimate that Orion's landing weight of 7337 kg would be cut in half by the removal of the chutes, the thermal backshell, the heat shield, and the heat shield carrier structure, then that gives us roughly 3700 kg. If you attached that to the "stock" Orion service module, it would have 2054 m/s of dV. Unfortunately the OMS engine stuck underneath is not powerful enough to get off the moon. And that leaves you with no budget for airlock, etc.