totm dec 2023 Artemis Discussion Thread

[tater]

1 hour ago, Exoscientist said:

 In the next post I’ll do the estimate with the four RL-10’s swapped out for the J-2X.

Other than an added time delay measured in years (10?), what do you think adding an engine 2.7X heavier, and lower I_sp will buy SLS?

The only useful thing EUS does, IMHO, is it improves launch scheduling because it can launch into a  circular orbit, then do phasing in LEO for TLI—vs ISPS which has to launch into the right trajectory from liftoff, giving us all the scheduling issues we saw with the first launch (important with FTS batter replacement needing a ride back to the barn and limited VAB<—>pad round trips).

Honestly, the most useful add-on to SLS to figure out? Moving the FTS batteries to someplace they can be swapped on the pad.

Edited December 18, 2023 by tater

[Exoscientist]

 Here’s an estimate of the SLS 1B payload to LEO with the RL-10’s swapped out for a J-2X.

For the dry mass of the Boeing EUS I increased the dry mass by 1,200 kg to account for the greater mass of the J-2X compared to the four RL-10’s. I kept the residuals for the engine at 3% though it may be less than this. Then the input data looks like this:

 

 And the results like this:

 

  About 119.7 tons to LEO. In the next post I’ll investigate increasing the payload by using a longer extension nozzle on the J-2X.

  Robert Clark

 

Edited December 20, 2023 by Exoscientist

[tater]

The claim for SLS to LEO with EUS is what, 105t? Their quoted payload to TLI (w/Orion) is 38t—ie: 11t comanifested with Orion.

For Block 2 they claim 130t to LEO, and 43t to TLI (w/Orion)—ie: 16t comanifested with Orion.

So still useless, even if you could get a Block 1B payload somewhere in between 1B and 2. Assuming Orion is not replaced (decades and billions), any change to SLS that doesn't buy ~70t to TLI is a waste of time.

 

[Exoscientist]

 There have been several stories recently coming out from NASA discussing the number of launches required for the NASA/SpaceX lunar landing. The undercurrent you sense is that NASA does not really like the plan:

Starship lunar lander missions to require nearly 20 launches, NASA says
Jeff Foust
November 17, 2023
https://spacenews.com/starship-lunar-lander-missions-to-require-nearly-20-launches-nasa-says/

And the reception that Destin of “Smarter Every Day” got in his critique of the plan before a room full of NASA engineers suggests he was saying out loud what NASA engineers couldn’t say.

 I proposed a single launch architecture using an Apollo-sized lunar lander with extra propellant being added to the Orions service module:

Possibilities for a single launch architecture of the Artemis missions, Page 2: using the Boeing Exploration Upper Stage.
http://exoscientist.blogspot.com/2023/08/possibilities-for-single-launch.html 

Because of the greater mass to be sent to TLI though, it may require greater payload capability than the SLS 1B’s 105 tons to LEO. So I was investigating methods of giving SLS greater payload capacity.

 The SLS 2 will get 130 tons to LEO but it won’t be here until the 2030’s and will require further billions to develop the upgraded SRB’s. 

 (If anyone tries to convince you the new SRB’s won’t cost much, please see my sig file.)

  Robert Clark

Edited December 20, 2023 by Exoscientist

[tater]

Single launch architectures with small landers are not "Artemis," though.

[Exoscientist]

On 12/7/2023 at 3:14 PM, sevenperforce said:

Look at the RL10A-4-2. It has an expansion ratio of 84:1, lower than the expansion ratio of the J-2X, but it achieves 451 seconds I_sp because it is an intrinsically higher-performing engine. Gas generator engines simply do not have the same efficiency potential as expander cycle engines.

 

The comparison between the RL10A-4 and the J-2X is illustrative. The RL10A-4 has an expansion ratio 84 to 1 and gets an Isp of 451 s. The J-2X has an expansion ratio of 80 to 1 and gets an Isp of 448 s: http://www.astronautix.com/j/j-2x.html . So the J-2X Isp is over 99% of the RL10A-4 value.

 It is correct in general a gas generator engine loses efficiency compared to an expander cycle engine because the gas generator dumps some of the burned gases overboard. But another consideration is combustion chamber pressure. Higher chamber pressure allows higher Isp. The chamber pressure of the RL10’s is about ~40 bar while for the J-2X, it’s about ~100 bar. So if the same close Isp value holds for the ultra high expansion ratio case then the modeled 482.5 s Isp for an expander cycle engine would only be reduced to ~479.3 s for the J-2X.

 But you do have to do trades on whether the increased weight is worth the increased payload permitted by the higher Isp. I’ll estimate the increased weight by what happened with the RL10. From the table below adding the longer nozzle extension to the RL10-A4 to get the RL10-B2 added about 110 kg to the engine mass, while increasing the Isp from 451 s to 465.5 s.

 The J-2X is a 13 times larger engine as measured by thrust, so I’ll estimate the nozzle extension as 13 times heavier so about 1,400 kg. But we might be able to make the added mass somewhat smaller. The current nozzle extension on the J-2X is metallic. We can reduce the mass of the current nozzle extension by a half or more by using carbon fiber.

RL10B-2

Active

1998

277 kg (611 lb)

110.1 kN (24,800 lbf)

465.5 s (4.565 km/s)

2.2 m (7 ft 2 in) Extended: 4.15 m (13 ft 7.5 in)

2.15 m (7 ft 1 in)

40:1

5.88:1

280:1

44.12 bar (4,412 kPa)

5-m: 1,125 s 4-m: 700 s

Delta Cryogenic Second Stage, Interim Cyrogenic Propulsion Stage

[1][42]

RL10A-4-1

Retired

2000

167 kg (368 lb)

99.1 kN (22,300 lbf)

451 s (4.42 km/s)

1.78 m (5 ft 10 in)

1.53 m (5 ft 0 in)

61:1

84:1

42 bar (4,200 kPa)

740 s

Centaur IIIA

[11][43]

RL10A-4-2

Active

2002

168 kg (370 lb)

99.1 kN (22,300 lbf)

451 s (4.42 km/s)

1.78 m (5 ft 10 in)

1.17 m (3 ft 10 in)

61:1

84:1

42 bar (4,200 kPa)

740 s

Centaur IIIB Centaur SEC Centaur DEC

[11][44][45]

  For a nozzle extension to 465.5 s on the J-2X and adding on also the 1,400 kg increased engine mass this would increase the LEO payload from 119.7 tons to ~122 tons by the Silverbirdastronautics.com estimator. This small size increase would be of doubtful value.

 If the 482.5 s nozzle extension could be added without much greater mass than the 1,400 kg, when taking into account also the current metallic nozzle could be reduced by carbon fiber, then we could get ~128 tons to LEO.

  Robert Clark 

Edited December 20, 2023 by Exoscientist

[tater]

6 hours ago, Exoscientist said:

 If the 482.5 s nozzle extension could be added without much greater mass than the 1,400 kg, when taking into account also the current metallic nozzle could be reduced by carbon fiber, then we could get ~128 tons to LEO.

Still not enough. That's less than Block 2, and Block 2 is also useless.

Targets for meaningful SLS changes are:

1. Mass to orbit: ~70t to TLI (or better), or don't bother.

2. Cost reduction. Think of a target in the 1 order of magnitude less expensive range given SLS capabilities. A few hundred million for the entire stack minus payload cost or it's not worth it.

3. Changes that improve launch schedule risk. This is low hanging fruit. EUS and making FTS battery swaps a pad thing would seriously move the dial.

Note that if #2 were to happen, then all sorts of changes start looking OK that are a waste of time right now. Just need to get a marginal launch from $4B to $400M.

Edited December 20, 2023 by tater

[sevenperforce]

6 hours ago, Exoscientist said:

On 12/7/2023 at 3:14 PM, sevenperforce said:

Look at the RL10A-4-2. It has an expansion ratio of 84:1, lower than the expansion ratio of the J-2X, but it achieves 451 seconds I_sp because it is an intrinsically higher-performing engine. Gas generator engines simply do not have the same efficiency potential as expander cycle engines.

The comparison between the RL10A-4 and the J-2X is illustrative. The RL10A-4 has an expansion ratio 84 to 1 and gets an Isp of 451 s. The J-2X has an expansion ratio of 80 to 1 and gets an Isp of 448 s: http://www.astronautix.com/j/j-2x.html . So the J-2X Isp is over 99% of the RL10A-4 value.

I'm not sure where astronautix got the 80:1 figure, but the actual manufacturer said the expansion ratio was 92:1. So I'm inclined to believe them.

Which goes to my point. You can't squeeze expander cycle performance out of a gas generator engine. It's a different beast.

[tater]

1 minute ago, sevenperforce said:

I'm not sure where astronautix got the 80:1 figure, but the actual manufacturer said the expansion ratio was 92:1. So I'm inclined to believe them.

Which goes to my point. You can't squeeze expander cycle performance out of a gas generator engine. It's a different beast.

The underlying problem with SLS was that it was designed with no mission in mind to set the minimum performance characteristics. So I'd say that any discussion about "fixing" the vehicle needs to define the mission goals, then look at changes that allow for that mission goal to be accomplished.

If the mission goal includes the idea of a 1-stack CSM/lander, then either Orion has to go, or it must send ~70t to TLI. Any other changes just add cost and schedule delay, for no gain at all.

If the goal is to send Orion to a distant lunar orbit (NRHO/Gateway), then the only gain past the already approved EUS is a few tons of extra comanifested payload mass to Gateway. Any change to the already established interstage for those payloads means a new MLP (>3 years and a billion bucks+?). Deciding if that is worth it requires a real payload example that would be nice to send. If it's not worth it, the payload is likely volume (not mass) constrained anyway.

If EUS can only get 38t to TLI, that implies the stack is ~107+ t in LEO. If all the extra 11t was in fact residuals for the EUS, the stack would have enough to put the whole thing in LLO with ~200 m/s margin. The EUS would need some ACES work done to mitigate boil off. That at least gives a mission architecture to "not Gateway." Not seeing any way to up the payload of SLS that does any more than what EUS might be able to do.

[SunlitZelkova]

@Exoscientist If you are really desperate to make a single stack launch with SLS, have you considered dropping Orion and having a CSM more akin to Gemini?

I feel like cutting out two crew would make the mass limitations more manageable. It’s just an idea that popped into my mind.

But, as @tater says, if you do that it wouldn’t be Artemis (that is, what NASA is trying to achieve with Artemis). Artemis is about going to the Moon to stay, and that requires a reusable lander- no matter the cost.

Speaking of reusability and high numbers of tanker flights, it should be noted during the planning for the original Space Transportation System proposal, the Reusable Nuclear Shuttle, which would take astronauts back to the Moon in the 1980s, would have required six Space Shuttle flights for refueling each time.

I don’t know if it is a good or bad thing to have a similar characteristic to the ill-fated Integrated Program Plan (which encompassed STS, a lunar base, and had a Mars mission on the end too).

I personally don’t think it is time to panic yet by the way. If Starship/Super Heavy gets to five* integrated flights with no successful orbit, I might start getting concerned, but it is still early in the testing phase.

And even if Starship doesn’t work great, SLS is certainly not the answer for a replacement for the reasons tater and co have said.

Exoscientist, how do you propose to lower the cost of SLS to make multiple lunar missions a year feasible?

*I chose this number because this is the number of flights it probably would have taken to get a successful N1 launch. N1 might have succeeded on the fifth but got cancelled after the fourth failed launch.

[DAL59]

https://www.nasa.gov/news-release/nasa-shares-progress-toward-early-artemis-moon-missions-with-crew/

Artemis II delayed 10 months to September 2025
III delayed to September 2026
Artemis I was six years behind schedule, so this is actually above average performance for this "program"

[Minmus Taster]

2 hours ago, DAL59 said:

https://www.nasa.gov/news-release/nasa-shares-progress-toward-early-artemis-moon-missions-with-crew/

Artemis II delayed 10 months to September 2025
III delayed to September 2026
Artemis I was six years behind schedule, so this is actually above average performance for this "program"

[mikegarrison]

No real surprise here.

In case people didn't notice, the Dec 2024 announced schedule was timed so that it would take place during the (presumed) second term of the President who announced it. Presidents usually aren't interested in funding PR opportunities for their successors.

[tater]

24 minutes ago, mikegarrison said:

In case people didn't notice, the Dec 2024 announced schedule was timed so that it would take place during the (presumed) second term of the President who announced it. Presidents usually aren't interested in funding PR opportunities for their successors.

To be fair, the date was likely an internal sales pitch from Bridenstine, who I think was an unambiguously good NASA Admin. His goal being to create urgency so the program would survive—which worked.

[mikegarrison]

12 minutes ago, tater said:

To be fair, the date was likely an internal sales pitch from Bridenstine, who I think was an unambiguously good NASA Admin. His goal being to create urgency so the program would survive—which worked.

When you are pitching to your boss, it's not great to lead with "We'll be going back to the moon! After you are out of office."

[tater]

6 minutes ago, mikegarrison said:

When you are pitching to your boss, it's not great to lead with "We'll be going back to the moon! After you are out of office."

The boss would have to be self-aware enough to realize that was a possibility

 

[Ultimate Steve]

Another thing I've thought up in another discussion. People often do a lot of Artemis to Apollo comparisons, and those not in the loop are often very confused by how goofy Starship looks next to Gateway. I think I've come up with some words that adequately sum up why.

Apollo was a program defined by what was possible to make in a decade. Artemis is a program defined by what we have available right now, as the political will to build anything dedicated from scratch is very low.

What we happen to have right now just so happens to be Beefy Apollo incapable of getting a lander to Low Lunar Orbit, low flight rate orange rocket incapable of doing much more than launching Beefy Apollo, promising medium and medium-heavy commercial rockets, minimal lander budget, and ambitious aspiring Mars colony builder that is also capable of landing on the Moon if rapid refueling works. Plus Blue Moon but I couldn't figure out a funny name for it.

That is the gist of why Artemis looks so goofy at first glance. A tower built with a random fistful of Legos is gonna look a lot different than if you could select those Legos.

[sevenperforce]

6 hours ago, Ultimate Steve said:

A tower built with a random fistful of Legos is gonna look a lot different than if you could select those Legos.

I couldn't have said it better myself.

[Exoscientist]

On 1/4/2024 at 4:32 PM, sevenperforce said:

No one wants a sortie lander. To achieve the goals of Artemis, we need substantial downmass -- something closer to the 46 tonne (launch) mass of the Altair lander. Add a reasonably-sized crew capsule, and doing this in a single launch means a vehicle capable of throwing upwards of 70 tonnes to TLI.

 

 

 I’m responding to this comment from the Starship discussion  here since my response is more general in nature on lunar missions.

 I don’t agree you need rockets giving 70 tons to TLI to get sustainable architecture allowing manned lunar surface stations, a la how the ISS is in low Earth orbit. As I mentioned before the SLS is too expensive, upwards of $4 billion per flight, to be used for cargo only missions. Better to use far cheaper commercial flights for that purpose.

 Robert Zubrin gave a plan for producing a Moon base using three launches of Falcon Heavy plus a launch of the Falcon 9 to carry the crew to LEO:

Op-ed | Moon Direct: How to build a moonbase in four years
Robert Zubrin March 30, 2018

https://spacenews.com/op-ed-moon-direct-how-to-build-a-moonbase-in-four-years/

 Key for his plan is using a “Lunar Excursion Vehicle” (LEV) of ca. 12 ton gross mass that is hydrolox powered as a lunar lander. This would require near zero-boiloff tech, but Zubrin thinks this is doable with current tech.

 For the two cargo missions launched by Falcon Heavy, Zubrin would use a larger hydrolox stage for the lander at ca. 40 ton size, a bit smaller than Centaur V, that could deliver ca. 12 tons of cargo to the lunar surface.

 Quite notable about his plan is as far as the manned flights it would require a single Falcon Heavy rocket as the launcher to get the hydrolox in-space stages to LEO. The FH has a payload capacity of 63 tons to LEO. So an only  63 ton launcher could get the required in-space stages to LEO, which could then do a manned round trip flight to the lunar surface. This small size for the launcher is coming from the fact the in-space stages are so much lighter being powered by hydrolox.

 The Falcon Heavy however is not manrated though. So it would need a separate man-rated launcher to get the astronauts to LEO. This man-rated launcher could be the Falcon 9 but it doesn’t have to be.

  Robert Clark

Edited January 11 by Exoscientist

[tater]

13 minutes ago, Exoscientist said:

 I don’t agree you need rockets giving 70 tons to TLI to get sustainable architecture allowing manned lunar surface stations

He said if you use a single stack architecture you need ~70t. This is because you need >2 crew on the surface for 1-2 weeks as a minimum mission profile. We've all been pitching distributed launch lunar missions here for years.

16 minutes ago, Exoscientist said:

 Key for his plan is using a “lunar exclusion vehicle” (LEV) of ca. 12 ton gross mass that is hydrolox powered as a lunar lander. This would require near zero-boiloff tech, but Zubrin thinks this is doable with current tech.

You have to shoehorn whatever the lander is into a 145m³ fairing.

My take (years ago here) was use the fairing AS the vehicle.

Old enough that at the time I used a Dragon 1 pressure vessel graphic. Diagonal line is a fold down leg, flush on launch.

The capsule volume there is ~20m³. So we have <120m³ to play with assuming some slop.

[tater]

Note that FH is supposed to be getting a ~200m3 fairing for the Space Force, so that's on the table as well.

Note for clarity, I do not in fact mean using the carbon fiber fairing AS the vehicle, I meant to use the identical mold line, since the forces on the vehicle would be well understood by SpaceX.

So in this case the aeroshell around the Dragon pressure vessel and removable nose cone would be fairing shaped. It would seamlessly transition to the mold line of any cargo/tankage, etc. The legs flush to fairing mold line, etc. Actual tankage slightly smaller dia. Assuming you need every cubic cm of volume. If you can et away with a smaller vehicle, then it can have the fairing separate normally.

[darthgently]

38 minutes ago, tater said:

Note that FH is supposed to be getting a ~200m3 fairing for the Space Force, so that's on the table as well.

Note for clarity, I do not in fact mean using the carbon fiber fairing AS the vehicle, I meant to use the identical mold line, since the forces on the vehicle would be well understood by SpaceX.

So in this case the aeroshell around the Dragon pressure vessel and removable nose cone would be fairing shaped. It would seamlessly transition to the mold line of any cargo/tankage, etc. The legs flush to fairing mold line, etc. Actual tankage slightly smaller dia. Assuming you need every cubic cm of volume. If you can et away with a smaller vehicle, then it can have the fairing separate normally.

I like this (but am out of "likes").  So while it would need to be human rated for TLI and lunar landing/ascent, if a normal Dragon with crew met it in a parking orbit, launch grade human rating wouldn't be necessary and could perhaps be coordinated with a prior crew-not-present propellent refill.  But being based on Dragon, it would logically have no huge barrier beyond paper and tests to becoming fully launch to lunar ascent human rated.  Cool idea

 

[tater]

2 minutes ago, darthgently said:

I like this (but am out of "likes").  So while it would need to be human rated for TLI and lunar landing/ascent, if a normal Dragon with crew met it in a parking orbit, launch grade human rating wouldn't be necessary and could perhaps be coordinated with a prior crew-not-present propellent refill.  But being based on Dragon, it would logically have no huge barrier beyond paper and tests to becoming fully launch to lunar ascent human rated.  Cool idea

Yeah, this was part of the idea when we were discussing it. Normal crew Dragon delivers crew. At the time we were also assuming storable props, so lower I_sp.

[sevenperforce]

13 minutes ago, Exoscientist said:

On 1/4/2024 at 4:32 PM, sevenperforce said:

On 1/3/2024 at 12:06 PM, Exoscientist said:

My opinion a moon rocket should be A moon rocket(singular). We did this 50 years ago. There is no reason why we can’t do that now.

No one wants a sortie lander. To achieve the goals of Artemis, we need substantial downmass -- something closer to the 46 tonne (launch) mass of the Altair lander. Add a reasonably-sized crew capsule, and doing this in a single launch means a vehicle capable of throwing upwards of 70 tonnes to TLI.

I don’t agree you need rockets giving 70 tons to TLI to get sustainable architecture allowing manned lunar surface stations, a la how the ISS is in low Earth orbit. As I mentioned before the SLS is too expensive, upwards of $4 billion per flight, to be used for cargo only missions. Better to use far cheaper commercial flights for that purpose.

 Robert Zubrin gave a plan for producing a Moon base using three launches of Falcon Heavy plus a launch of the Falcon 9 to carry the crew to LEO...

If you go back and look at the actual sequence of replies, you'll see that I was replying to your comment where you said that a moon rocket should be a single-launch affair: "We did this 50 years ago. There is no reason why we can't do that now." It's disingenuous to start by criticizing a distributed-launch architecture on the grounds that we need a single-launch architecture, then respond to criticisms of a single-launch architecture by saying we actually need distributed launch.

24 minutes ago, Exoscientist said:

I don’t agree you need rockets giving 70 tons to TLI to get sustainable architecture allowing manned lunar surface stations...

Zubrin's 2018 op-ed (which was very short on detail) proposed three Falcon Heavy launches to LEO at 60 tonnes each, in the form of a cargo lander that could take itself from LEO to the lunar surface. Assuming 453 seconds I_sp, that's 29.2 tonnes to TLI per launch, which comes to 87.6 tonnes to TLI total, substantially more than the 70 tonnes I proposed for a single-launch architecture that you asked for.

Don't get me wrong -- I love the idea of distributed launch. In fact, that was my point; I was explaining why a single-launch architecture was prohibitively difficult SPECIFICALLY because you said it was necessary. Distributed launch is definitely the way to go. That said, Zubrin's numbers are...rather aspirational. You need 5.93 km/s Δv to go in either direction between LEO and the lunar surface. Starting at 60 tonnes in LEO with engines pushing 453 s I_sp, you would need to burn 44.2 tonnes of hydrolox to land on the moon with bone-dry tanks. You'll need to pull this off in a single TLI burn to avoid Oberth losses and multiple passes through the Van Allen belts -- no less than 400 seconds preferably. So you need to push at least 110.5 kg of propellant through your engines per second, giving you a minimum thrust of 491 kN. The RL10C-1-1 gives you a thrust of 106 kN, so four won't be enough. But even with four of them in an EUS-style cluster, that's 752 kg of engine alone, leaving a mass budget of only 3 tonnes to hold over 44 tonnes of hydrolox. By comparison, Centaur III's tanks weigh in at just over 2 tonnes and carry less than 21 tonnes of hydrolox.

Such a vehicle wouldn't even fit on Falcon Heavy. Hydrolox has a bulk density of around 0.36 kg/L, meaning that ~45 tonnes will require 125 cubic meters of tank volume. Four RL10C-1-1s will fit inside the fairing -- barely -- but they take up 2.4 meters of vertical space (although a carefully-designed thrust plate can reduce this to around 2.13 meters). The Falcon Heavy extended fairing has a usable internal diameter of 4.57 meters, but we need room for landing legs. The landing legs on the Apollo Lunar Module added just over 2 meters to its diameter while stowed on top of the S-IVB, but let's assume implausibly that a modern design can cut this in half, requiring only 50 centimeters of stowed clearance per landing leg. Taking tank skin thickness to be negligible (not a good idea for a lander, mind you), that gives us a maximum internal tank radius of 1.785 meters, for a cross-sectional area of 10 m². Assuming ellipsoidally-capped tanks with a rule-of-thumb half-radius height, the total stage height would need to be 15.2 meters, leaving a cosy 1.34 meters of vertical volume on top and impinging on the fairing separation system:

Spoiler

Maybe we can get VERY short astronauts.

I'm not sure how Zubrin expects anyone to construct a moon lander slightly taller than the New Shepard booster, complete with landing legs and thrust structures and egress ladders, all while somehow getting a 44% higher mass ratio than Centaur III. 

[tater]

6 minutes ago, sevenperforce said:

meaning that ~45 tonnes will require 125 cubic meters of tank volume.

As I recall my suggestion (ages ago) with the diagram above was partially to use every single m³ of volume by making the tank the fairing. Even then the fairing size available was not adequate with a Dragon pressure vessel.

At the time I think that various F9/FH ideas were not predicated on a embark crew to lander at LEO, and fly propulsively back to LEO model. The dv requirements there are fairly prohibitive. Anyway, there are options for a distributed launch model.