sevenperforce

That's Oberth for you. The moon might not have much gravity but it definitely gives you an advantage. Ah, I see my mistake. I was following Wikipedia, which listed the heaviest AJ10 variant at 100 kg. Just took a look at astronautix and I see there is a huge array of different mass variants. Gonna have to run some new numbers to account for that. Since we are talking about liftoff thrust, it would be worth factoring in some of the information from here about ascent profiles. For example, the Apollo LM AV had a ten-second vertical burn and then a total pitchover. That corresponds to 16.2 m/s of gravity drag losses; if you have more thrust, you can reduce that.

There's a high sensitivity to things like mass ratio, too. Transtage had a 16.995% dry mass fraction (not counting the engine) while the Delta-K and Delta-P had a 12.4% dry mass fraction. So that makes a significant difference.

Not to fanboy but it really is like a bunch of ten year olds on a track team while Usain Bolt is running laps....

Were you using a single insertion burn or were you doing the powered flyby? I think that doing a single insertion burn at NRHO is more along the lines of 600-700 m/s. It's worth noting that the original NASA HLS RFP we were reviewing earlier definitely had a higher quoted mass for the ascent vehicle: 9-12 tonnes is significantly larger than the 6.5 tonnes that the stretched Blue Moon can land, so that's odd. Using a 9-12 tonne AV rather than the 6.5 presumed for the LockMart reusable AV, we'd be seeing a crew capsule that's 2.9-3.8 tonnes apart from engines and tankage. That's much better margin, but not SO much more that the 6.5-tonne version is utterly implausible. I certainly wouldn't be choosing an 11 kN engine. I was baselining with the AJ-118K, at 44 kN max thrust, hover at 25% throttle, 100 kg mass. A Shuttle OME would also do the trick, though not as spry at liftoff. Why do you need multiple gees of thrust during descent?

One final configuration that may prove to be the most efficient of all: you can build the LAM with a single, smaller drop tank, enough to give it around 1400 m/s of dV...around 4.3 tonnes of propellant. Wet mass to LEO is just under ten tonnes. You then pair it in LEO with a Centaur or FHUS, which pushes it to TLI and performs the powered lunar flyby and NRHO insertion. The manned crew taxi is fueled and mated, then that same high-performance upper stage performs the transfer to LLO and initiates the first 150 m/s of descent before breaking free. This is probably the most mass-efficient approach, though I would have to run the numbers. It does require five restarts of the upper-stage engine. However, it's also extremely safe. If for some reason the LAM engine fails after the initiation of descent or the upper stage fails to separate from the LAM, the crew taxi has ample dV and plenty of time to separate, cancel descent, and return to LOP-G.

They have. They were the first country to sign-on. Japan was the second. We're still waiting on the ESA and Roscosmos, but both are almost guaranteed to participate. Yes, a WHILE back, but NASA's most recent request for submissions for a HLS stated that a Canadarm would not be installed at LOP-G initially.

By my numbers, the mass of the cabin and associated structure for the Apollo LM AV (no prop tanks or engine) was 1,600 kg, plus crew and cargo. Mass growth of only 30% (for the LockMart reusable AV concept) seems pretty minimal.

This is one of the bridges; we would need a Canadarm added to LOP-G. Though all it has to do is move a single replaceable tank from a mount on the ascent vehicle and attach it to the reusable crew taxi while both are docked. No major assembly. Well, to begin with, the cost to get from TLI to LOP-G is not 800 m/s; that's the round-trip cost for Orion. If you are making a one-way trip from TLI to LOP-G, you only need a 183 m/s powered lunar flyby and a 215 m/s NRHO insertion burn, for a total of ~400 m/s. See page 5 of this paper. I was allowing 430 m/s as this is the value cited in some other papers. But in any event, there are multiple ways to do it: Launch the landing/ascent module (LAM) as a monolith to TLI with drop tanks so it can use its AJ-10 to do the powered flyby and NRHO insertion. In this configuration, the LAM needs two drop tanks, one for the trip to LOP-G and one for the trip to the lunar surface, and masses around 20 tonnes. You would need distributed launch, with the monolith going to LEO on one commercial launch vehicle and mating to a separately-launched docking ring on a Centaur or FHUS that would push it to TLI. You can also launch the LAM with only one drop tank (for the trip from LOP-G to the lunar surface) and keep it attached to the TLI stage so that the TLI stage provides the 400 m/s required for the trip to NRHO. With this config, the monolithic LAM masses 17-18 tonnes so you would still need distributed launch, but you could probably do it a lot cheaper because the TLI stage will be more efficient. Finally, you can launch the LAM with no drop tank at all and use the TLI stage (maybe ACES?) for NRHO insertion AND transfer to LLO and finally to the lunar surface, but that requires a lot of restarts. This is by far the most efficient, though. The LAM masses under 7.6 tonnes. Two things. First, your AV mass seems high. As @tater points out, the stretched Blue Moon lander can put no more than 6.5 tonnes on the lunar surface, so that's the upper limit on LockMart's reusable AV wet mass. With hypergols, it needs to burn 56% of its weight to go from the lunar surface to LOP-G. Factoring in dry mass for the AJ-10 and tankage, the LockMart AV has only 2,070 kg for capsule, LS, crew, return cargo, airlock, and everything else you mentioned. Not a lot, but keep in mind that the dry mass of the Apollo LM AV was only 2,050 kg including tanks, so this is reasonable. Additionally, I'm dispensing with the descent vehicle altogether. Wet mass on the surface is under 7 tonnes and is merely the reusable crew taxi, the ascent motor+tanks, and landing legs/ladders that are jettisoned at takeoff. The descent phase is either performed by the transfer vehicle as a crasher stage, or it is performed using drop tanks on the ascent vehicle. It makes no sense to have a large landing structure to support dry mass you no longer need. EDIT: To @tater's point, here's the quote: From here.

All things being equal, this is the easy part. If you are producing the liquid methane and LOX, you have the capacity to condense, so you have the capacity to re-condense boil-off.

I didn't listen to the audio (out in a place where I can't) but I see 6 RL-10s on the descent stage and 4 engines (maybe RL-10s, but probably AJ-10s) on the ascent stage?

I think this may become my most-liked post...........

On a final note -- I have not yet done the math, but I think it is very close whether an expendable Falcon Heavy could reach LOP-G with the FHUS still attached to the LAM and still have enough kerolox reserves to act as the crasher stage from LOP-G through LLO to the lunar surface. An expendable Falcon Heavy with crossfeed could probably pull it off.

You're hired!!

I did leave a little bit of optimization on the table with this analysis, based on drop tank size. The size of the drop tank on the crew taxi is constrained -- you don't want to give it much more dV responsibility or you end up with a potential thrust shortfall -- but making the "built-in" tankage on the LAM exactly four times more was arbitrary. I could have made them bigger, which means they take over earlier in the descent, or smaller, which means they would do less of the descent and essentially take over just before landing. Probably best to make them as small as possible, but getting rid of the dry mass of the transfer tanks early is important too. If the ~20-tonne stack is launched monolithically into LEO, all the drop tank couplings can be performed on the ground, which is obviously preferred. It is hard enough to design drop tanks without having to worry about attaching them on orbit. The drop tank on the crew taxi would of course need to be attached at LOP-G, but since it is not removed until the return, it isn't a "drop" tank in the same way. The tanks would have valves to determine drain priority, which is also straightforward enough. The iterated-drop-tank design has fairly good abort characteristics; the crew taxi can separate and return to LOP-G if there is a problem up through LLO, and can correct a serious thrust/fuel anomaly on ascent by completing part of the ascent on thrusters alone (though it would then need a rescue mission). The LAM+taxi architecture would benefit from using RD-4 thrusters, which boast higher specific impulse and higher thrust than the Dracos. The extra 4.5% in Isp for the taxi element reduces mass by a few percent all the way down the line. It could also be used in conjunction with a cryogenic transfer stage. If the LAM alone was launched onto TLI, it would need only a small drop tank to brake itself into rendezvous with LOP-G -- almost exactly the size of the one on the crew taxi -- and total mass to TLI would be just 6,473 kg! Towing that (minus rendezvous drop tank, plus the crew taxi) from LOP-G to within 300 m/s of the lunar surface is still 2,304 m/s, but with cryogens it gets more interesting. Using the RL-10, for example, you only need about 5.5 tonnes of hydrolox in a balloon tank. Of course, that transfer stage needs to get itself to LOP-G, which means it needs RCS thrusters and another 430 m/s of dV, but you only have to throw about 7-8 tonnes to TLI.

Upon review of the drop tank architecture...if we are doing hypergolics, there is no need for multiple engines. One will do. A downthrottled vacuum-expanded Superdraco can get no lower than 12 tonnes lunar thrust, so it is just too thrusty. A deep-throttling AJ-10 or comparable engine will need to be used instead. @tater gives the BlueMoon+LockMart ascent vehicle a dry mass of 2.5 tonnes, but I'll be a touch more generous and say 2.8 tonnes (assuming 319 s Isp, 2.6 km/s from lunar surface to LOP-G, 6.5 tonnes on the surface). Those 3.7 tonnes of props will require a tankage dry mass of 630 kg and of course 100 kg of AJ-10 engine, leaving the crew vehicle plus control thrusters at a mass of about 2,070 kg. So if that's our goal, let's see what the reusable-crew-taxi-with-drop-tanks approach will get us. Let's say it has Draco thrusters at 300 s Isp and performs the last 200 m/s of LLO circularization after the lander/ascent module ("LAM") is jettisoned. It needs a dV of 930 m/s to circularize and return to LOP-G. It needs 822.3 kg of props and with 139.7 kg of tankage, so our drop tank will mass 962 kg and our "crew taxi" will mass 3,032 kg at LAM separation. Let's throw on 100 kg for the AJ-10 at 319 s. If we bolt four of those same expendable tanks around the engine on the lander/ascender, we end up with 1.992 km/s on the vehicle, which gets you into LLO easily. Since you only need 1.67 km/s (since the crew taxi does the circularization), you only need 2,606 kg of props for ascent, leaving you with 683 kg of props for final approach and landing. Of course, on the surface, you're going to need to support the landed mass of 6,297 kg. Let's add the 9.4% structural fraction for landing legs, ladder, and associated compressive structure that will be jettisoned at takeoff (landing mass is 6,889 kg, which can be hovered if you downthrottle to 11.1 kN, or about 25% of max thrust on an AJ-10). That additional 592 kg gives us a lander/ascender dry mass (including the four bolted tanks) of 1,251 kg and means our 683 kg of propellants allow us to handle the last 296 m/s of descent and landing. Accordingly, our transfer vehicle would need to provide 2,304 m/s from LOP-G to the lunar surface. If we use the lander/ascender's AJ-10 to accomplish that with drop tanks, we will need larger drop tanks with a total of 10.13 tonnes of propellant and 1.72 tonnes of tankage. Remember, however, that the crew taxi is reused every flight, so that's 2,070 kg you don't have to throw to TLI or brake into TLI. With this caveat, the total delivery to LOP-G for each completed sortie becomes 17.35 tonnes. Rendezvous of THAT to LOP-G from TLI requires another 430 m/s, which is 2,624 kg of propellant and 446 kg of drop tankage. Total mass sent to TLI: 20.42 tonnes. You can easily put 20.42 tonnes into LEO with a triple-core-recovery Falcon Heavy and then send up an expendable Falcon Heavy with a naked upper stage and a docking ring to throw that into TLI.

Any indication on how the LockMart Ascent stage is supposed to be reused? Are they going to refuel with prop transfer at LOP-G? If they went with a pump-fed engine they could save enough in tankage dry mass to afford two engines, which gives them engine-out capability.

Jupiter 120 plus Orion (which is the same as a two-engine NLS-3) would have been cheaper per launch than the Shuttle for crew transport to the ISS. Same SRBs, same tank, same propellants. Sacrifice 2 engines for the price of refurbishing 3, sacrifice one Orion for the price of refurbishing an entire Shuttle. No-brainer. Way more expensive than Falcon 9 + Dragon 2, of course. DIRECT intended to use the core configuration as a workhorse to fulfill as many roles as possible. Baseline configuration (external tank, 2-3 SSMEs, 4-segment SRBs), takes Orion to the ISS at lower cost than a Shuttle launch. Add another engine (still < Shuttle cost), and you can comanifest cargo with Orion a la STS. Add an upper stage, and you can send Orion BLEO or cargo. Upgrade to 5-segment SRBs and a fourth/fifth SSME, and you can comanifest cargo with Orion BLEO. In contrast, Congress and NASA batted SLS back and forth for so long that they never actually built it (which is still true) and so it could never be a workhorse, and so they had to make it more capable to do Special and Exciting things, so they had to stretch the tank and baseline at 5-segment SRBs and everything else.

Except for the part about the Shuttle not mitigating crew risk. I would rather fly on Orion on top of a Proton or Ares V than fly on the Shuttle.

There's something to be said for the presumed simplicity of a 1.5-stage-to-orbit vehicle. You can light everything on the ground, make sure it all checks out, and then launch. Same core for everything and make it modular. IF you have a regular need for high-mass launches (and that's the problem), you get nice economies of scale. NLS-1 would have been a four-engine version of Jupiter-130, with two four-segment SRBs strapped to an external tank with four SSMEs under it and an interstage+payload on top, placing 22 tonnes in LEO. NLS-2 would have been a Shuttle-derived version of the Saturn V-B concept or the Atlas LV-3B, with two sustainer SSMEs on the external tank and four SSMEs on a jettisonable skirt, placing 6 tonnes in LEO.

I only just recently realized that SLS was proposed even before Jupiter-DIRECT, as the "National Launch System" under Bush Sr. It was to be exactly the same plan as Jupiter-120 and Jupiter-130: slap a few used SSMEs (or cheaper, expendable STMEs) onto the back-end of a Space Shuttle main tank, drop an interstage on top, and go. They also had a plan to put a single STME on a 5-meter tank and add an RL-10 based upper stage, which of course is essentially what became the Delta IV. It was supposed to fly concurrently with the STS program. Clinton canceled it because they wanted to focus on SSTOs and increasing STS cadence.

Yeah, because autonomous docking and assembly around the Moon is easier and less risky than LEO (this isn't a poke at you, every time EoR comes up WRT the Moon and Artemis in here rendezvous and docking is somehow hard, but around the Moon it's not. For reasons.) Right. Plus, Earth orbit assembly allows a single TLI burn, a single lunar powered flyby, and a single LOP-G rendezvous. So it's more mass-efficient by definition, because you have less dry mass going to LOP-G. If you build the tanks larger, then that's more dry mass wasted at LOP-G, and conceivably more dry mass being hauled down to LLO and even further.

Survival on the lunar surface is supposed to run through the lunar night...seemingly multiple cycles, so I expect boil off is low enough for at least a few months, yeah. The ascent stage is supposed to be reusable. I would think they send the descent stage first, then the crew, then the transfer stage carrying the propellant to refuel the ascent stage. Ordering of the last two is optional.

...and a distributed launch scheduling nightmare. The New Glenn user's guide mentions no RCS on the booster but mentions four triaxial RCS thrusters on the upper stage for settling. Some triaxial thrusters are visible here, though this may be artistic license. Pressurization of both stages is confirmed to be autogenous. In the Blue Moon reveal video, we see four triaxial thrusters on the top of the lander and four quintaxial thrusters on the bottom. Bezos discussed that they will use hydrogen boil-off to chill the LOX tanks and then pump the warmed GH2 into accumulators, along with boiled GOX that goes into its own accumulators. The GH2+GOX will be used to operate fuel cells, as in the ACES design. However, I see no other tanks on the vehicle than the accumulators and the main tanks. If I had to make a wild guess, I would say that the 32 thrusters on Blue Moon are pressure-fed from the accumulators with spark ignition. Alternately, they could be cold-gas thrusters pressure-fed solely from the GH2 accumulators; this avoids the need for ignition altogether. These thrusters can be used for mid-course correction, reducing the impact of boil-off considerably.

From https://spaceflightnow.com/2019/10/22/bezos-says-space-industry-stalwarts-will-help-blue-origin-build-moon-lander/ Assumed the NG contribution was a version of Cygnus. Apparently not. So, Altair, but with a smaller ascent module and the BE-7 in place of the RL-10.