SpaceX to NASA quote : simplified mission architecture : Technical discussion

[DanClemmensen]

SpaceX is already developing a Ship very similar to Depot but with crew accomodations. It's called Starship HLS. If you want your mission to include a lifeboat, use another HLS.

HLS crew quarters are much larger, and reduce potential fuel tank size. Would the HLS have enough fuel capacity to refuel the landing HLS and retain enough fuel for both to return to LEO? Also HLS has landing thrusters, cargo space, not desired on a Depot. My argument is that the Dragon is already designed, and in production, building a whole other HLS would be more costly. The Dragon would be stripped of Draco and Super Draco thrusters.

Any new design will take time and development money. Another instance of an existing design will in general be cheaper and will almost certainly be quicker. If you want a lifeboat, send a lifeboat. If you want to refill, send a Depot. If you want both, send both.

[Vultur]

Yes, as a practical matter crewed in-orbit refill almost certainly has a very low pLOC, much lower than a lot of stuff NASA tolerates in SLS/Orion, and the formal pLOC will drop rapidly as experience accumulates with uncrewed refill.  Let's wait and fight that battle later.

OTOH, this may be exactly the time to fight that battle, now that the idea of reconsidering the plan of record has been mooted. An all-SpaceX architecture means no crew launch on solids; I doubt the risk of crewed refueling is at all comparable to that.

[DanClemmensen]

Yes, as a practical matter crewed in-orbit refill almost certainly has a very low pLOC, much lower than a lot of stuff NASA tolerates in SLS/Orion, and the formal pLOC will drop rapidly as experience accumulates with uncrewed refill.  Let's wait and fight that battle later.

OTOH, this may be exactly the time to fight that battle, now that the idea of reconsidering the plan of record has been mooted. An all-SpaceX architecture means no crew launch on solids; I doubt the risk of crewed refueling is at all comparable to that.

Maybe. both architectures are all-SpaceX. One has no refill on a crewed Ship, the other is cheaper but required crewed refill. Do the one that appears to be safer to kill that objection now. shift to the other one later after getting more experience.

[thespacecow]

What modifications would be required for D2 to handle deep space? Supposedly the heat shield is already designed for higher energy reentry. Is it communications that's the issue?

Yes, there are a number of possible issues that might require minor (or major) modifications to D2 if it were to be used for the trans-lunar and trans-Earth segments of the mission. Those include as you mention its ability to function in deep space and its ability to re-enter safely after a return from the Moon. The comms issue you mention is just one of several related to D2 electronics. The other would be radiation hardening. Finally, the mission architectures discussed require D2 to ride "on the nose" of the Starship-based transport vehicle, which in turn requires D2 to handle "eyeballs-out" acceleration.

None of these capabilities are proven. Even an aggressive program to test them would push the crewed mission outside the timeframe being discussed here.

I disagree, the timeframe being discussed here is 3 to 4 years, which is a long time for SpaceX development program. Remember it took only 2.5 years from the announcement of Polaris Dawn to its flight, and most of the development time in that case was spent on spacesuits.

3 to 4 years should be more than enough to modify Crew Dragon for lunar operations. BTW, current D2 can already handle "eyeballs-out" acceleration since that's how it does high delta-v maneuvers: by firing the thrusters around the docking port.

The main obstacle for using Crew Dragon in lunar operation is Elon Musk's desire to retire it asap and replace it with Starship. But I think old space p*ssed him off enough that he would be willing to accept a less than perfect solution in this case.

[mikelepage]

If there's a genuine simplification, I can't think of what it would be.  Everything requires refueling.  Everything requires landing a tall (maybe not quite as tall) Starship on the Moon.

Tried the simple mission I had in mind in Excel and my numbers seem to close (see spreadsheet)... This involves sending an HLS + tanker to the moon. HLS descends to lunar surface. HLS ascent from lunar surface to LLO is followed by single prop transfer in LLO (~400ton), followed by both performing TEI. HLS brakes propulsively into LEO, while tanker aerocaptures. D2 shuttles back and forth from LEO. What am I missing?

NASA has always talked about lunar exploration being a stepping-stone to Mars, and although that's not true in many ways, the one way it is true and relevant to SpaceX is that it's an opportunity to do *many* re-entry tests of Starship's heat shield at interplanetary velocities. Not with anyone onboard of course.

It seems to me that SpaceX would say using D2 as a lifeboat has about as much merit as designing a launch escape system for starship. Better to put more effort into making the main ship reliable.

That said, Earth EDL is the most intense challenge for starship's heat shield, so keep D2 as a proven Earth to LEO shuttle until crewed/P2P starship proves itself.

[TheRadicalModerate]

If there's a genuine simplification, I can't think of what it would be.  Everything requires refueling.  Everything requires landing a tall (maybe not quite as tall) Starship on the Moon.

Tried the simple mission I had in mind in Excel and my numbers seem to close (see spreadsheet)... This involves sending an HLS + tanker to the moon. HLS descends to lunar surface. HLS ascent from lunar surface to LLO is followed by single prop transfer in LLO (~400ton), followed by both performing TEI. HLS brakes propulsively into LEO, while tanker aerocaptures. D2 shuttles back and forth from LEO. What am I missing?

That's exactly the single-HLS, post-ascent refueling case.  Yes, it works easily, from a prop standpoint.  But it requires refueling in lunar orbit, after the HLS no longer has the prop to make it home if the refueling fails.  The refueling also has to happen with crew aboard, which maybe doesn't matter in this case, since the crew will die if the refueling fails from a simple docking problem.  An explosion would just make them dead slightly sooner.

It's also not a simplification.  Option A only requires one refueling in an HEEO that's reachable directly from launch.  After that, it heads off to NRHO to wait for the Orion.

I would consider either a major operational simplification or a substantial development simplification worthwhile.  But more RPODs and refuelings, even if they replace SLS/Orion, are more complex operationally.¹  And nothing we've discussed really reduces the developmental load:  Starship HLS still needs the same refueling, the same boiloff management, the same ECLSS and human factors work, the same landing legs and guidance, and the same landing thrusters, as the plan of record.

__________
¹The one absolutely true argument for SLS/Orion is that it's a much simpler conops than anything that could replace it quickly.  You launch it, you stabilize your staging orbit, and you go to TLI.  Then the Orion uses its AJ-10 (can't remember what they're calling it this time) to enter and leave NRHO.  Meanwhile, there are two RPODs between Orion and the HLS, and that's it.

That doesn't imply that it's safer, and it's certainly the greatest developmental trainwreck in NASA history.  But it's very simple.

[Legios]

What modifications would be required for D2 to handle deep space? Supposedly the heat shield is already designed for higher energy reentry. Is it communications that's the issue?

Yes, there are a number of possible issues that might require minor (or major) modifications to D2 if it were to be used for the trans-lunar and trans-Earth segments of the mission. Those include as you mention its ability to function in deep space and its ability to re-enter safely after a return from the Moon. The comms issue you mention is just one of several related to D2 electronics. The other would be radiation hardening. Finally, the mission architectures discussed require D2 to ride "on the nose" of the Starship-based transport vehicle, which in turn requires D2 to handle "eyeballs-out" acceleration.

None of these capabilities are proven. Even an aggressive program to test them would push the crewed mission outside the timeframe being discussed here.

I disagree, the timeframe being discussed here is 3 to 4 years, which is a long time for SpaceX development program. Remember it took only 2.5 years from the announcement of Polaris Dawn to its flight, and most of the development time in that case was spent on spacesuits.

3 to 4 years should be more than enough to modify Crew Dragon for lunar operations. BTW, current D2 can already handle "eyeballs-out" acceleration since that's how it does high delta-v maneuvers: by firing the thrusters around the docking port.

The main obstacle for using Crew Dragon in lunar operation is Elon Musk's desire to retire it asap and replace it with Starship. But I think old space p*ssed him off enough that he would be willing to accept a less than perfect solution in this case.

There's no reason to do this.  Orion and SLS already work and are proven.  There's no reason to modify a Dragon for deep space.  That part of the mission is solved.  It's the lander that needs work.

[greybeardengineer]

There's no reason to do this.  Orion and SLS already work and are proven.  There's no reason to modify a Dragon for deep space.  That part of the mission is solved.  It's the lander that needs work.

Orion has yet to fly with a functional ECLSS. NASA will actually flight test one with crew on board on an extended BEO trajectory. A fine example of the two tiered human rating system at the space agency.

[mikelepage]

Tried the simple mission I had in mind in Excel and my numbers seem to close (see spreadsheet)...

That's exactly the single-HLS, post-ascent refueling case.  Yes, it works easily, from a prop standpoint.  But it requires refueling in lunar orbit, after the HLS no longer has the prop to make it home if the refueling fails.  The refueling also has to happen with crew aboard, which maybe doesn't matter in this case, since the crew will die if the refueling fails from a simple docking problem.  An explosion would just make them dead slightly sooner.

Given that we don't get to this scenario unless SpaceX has already demonstrated prop transfer that works well enough in LEO, I'm not entirely sure why we think it's more risky in cis-lunar space? Maybe it's not where it is, per se, but how long the tanker will be waiting around in lunar orbit? I'm assuming it would have to go TLI before HLS, but it may not even be waiting around as long as a depot-refuelling campaign.

It's also not a simplification.  Option A only requires one refueling in an HEEO that's reachable directly from launch.  After that, it heads off to NRHO to wait for the Orion.

I just saw that Artemis 1 measured 27-35 mSv ionising radiation exposure, presumably from just two passes through the belt. Compare to this chart I generated some years back, showing yearly radiation exposure at various orbital altitudes/inclinations.

If transferring crew to HLS in HEEO only means one or two extra passes through the belts, it's probably justifiable. Still I'm not sure how the human factors team would rate the possibility of needing to refuel during an emergency mission abort, versus guaranteed doubling or more radiation exposure. 

I would consider either a major operational simplification or a substantial development simplification worthwhile.  But more RPODs and refuelings, even if they replace SLS/Orion, are more complex operationally.¹  And nothing we've discussed really reduces the developmental load:  Starship HLS still needs the same refueling, the same boiloff management, the same ECLSS and human factors work, the same landing legs and guidance, and the same landing thrusters, as the plan of record.
__________
¹The one absolutely true argument for SLS/Orion is that it's a much simpler conops than anything that could replace it quickly.  You launch it, you stabilize your staging orbit, and you go to TLI.  Then the Orion uses its AJ-10 (can't remember what they're calling it this time) to enter and leave NRHO.  Meanwhile, there are two RPODs between Orion and the HLS, and that's it.

That doesn't imply that it's safer, and it's certainly the greatest developmental trainwreck in NASA history.  But it's very simple.

Agreed, but I'd be inclined to go with the proposal that aims to make such operations a routine occurrence.  As someone who used to avidly watch every Falcon 9 launch, I couldn't even tell you how many have launched this last week, but I'm sure there's been more than one.

[Hug]

Alas, both links provided by Grok are either dead or not accessible by me. HLS Operations Concept Document (OCD) https://ntrs.nasa.gov/api/citations/20250004567/downloads/Artemis-HLS-OCD-Q3-2025-Redacted.pdf
and https://www.nasa.gov/artemis/library
and https://www.nasa.gov/artemis-iv

On re-reading, I made a bad interpretation. There are only discussions about post Artemis IV refueling of HLS, not actual demonstration.

I guess I get too wrapped up in SpaceX mantra of re-usability and cannot imagine a new HLS being built and discarded for each follow on moon mission beyond Artemis IV. But I suppose NASA thinks that way.

Those links are a hallucination by Grok and never existed.

It's SpaceX that thinks that way when they built a non-ZBO lander that to be reusable has to loiter for month-year long periods. A thing I've come to accept is that Starship is intending to be the most expendable and reusable rocket. Fundamentally they want to send hundreds (if not thousands) of them at Mars with no intention of retrieving; expending the stages at a higher rate than most expendable launch vehicles. Expend HLS's see how I care.

[woods170]

Alas, both links provided by Grok are either dead or not accessible by me. HLS Operations Concept Document (OCD) https://ntrs.nasa.gov/api/citations/20250004567/downloads/Artemis-HLS-OCD-Q3-2025-Redacted.pdf
and https://www.nasa.gov/artemis/library
and https://www.nasa.gov/artemis-iv

On re-reading, I made a bad interpretation. There are only discussions about post Artemis IV refueling of HLS, not actual demonstration.

I guess I get too wrapped up in SpaceX mantra of re-usability and cannot imagine a new HLS being built and discarded for each follow on moon mission beyond Artemis IV. But I suppose NASA thinks that way.

Those links are a hallucination by Grok and never existed.

Correct for all three links: they don't exist. Every single one resulting in a 404 error. And yet another fine example of why people, at this stage of development, should not blindly trust whatever BS is spewed by AI.
AI hallucination is a real problem and just one of many reasons why AI is not trustworthy at this time.

Edit: I pointed out the above some weeks ago in another thread.

[Blackhorse]

Grok hallucinating SpaceX links ? the irony-meter has gone off-scale.

[woods170]

Grok hallucinating SpaceX links ? the irony-meter has gone off-scale.

Well, none of the three non-existing links were "SpaceX links". All three of them were pointing to non-existent NASA content. What you have to remember here is that most NASA material on HLS is generic in nature, applying to both HLS contractors (SpaceX and Blue Origin).

[thespacecow]

I disagree, the timeframe being discussed here is 3 to 4 years, which is a long time for SpaceX development program. Remember it took only 2.5 years from the announcement of Polaris Dawn to its flight, and most of the development time in that case was spent on spacesuits.

3 to 4 years should be more than enough to modify Crew Dragon for lunar operations. BTW, current D2 can already handle "eyeballs-out" acceleration since that's how it does high delta-v maneuvers: by firing the thrusters around the docking port.

The main obstacle for using Crew Dragon in lunar operation is Elon Musk's desire to retire it asap and replace it with Starship. But I think old space p*ssed him off enough that he would be willing to accept a less than perfect solution in this case.

There's no reason to do this.  Orion and SLS already work and are proven.  There's no reason to modify a Dragon for deep space.  That part of the mission is solved.  It's the lander that needs work.

The main reason to do this is to kill off SLS/Orion.

But it just so happens using Dragon for deep space also simplify the architecture/conops (slightly) and improves crew safety (by a lot), and may be able to speed things up (slightly).

[DanClemmensen]

If there's a genuine simplification, I can't think of what it would be.  Everything requires refueling.  Everything requires landing a tall (maybe not quite as tall) Starship on the Moon.

Tried the simple mission I had in mind in Excel and my numbers seem to close (see spreadsheet)... This involves sending an HLS + tanker to the moon. HLS descends to lunar surface. HLS ascent from lunar surface to LLO is followed by single prop transfer in LLO (~400ton), followed by both performing TEI. HLS brakes propulsively into LEO, while tanker aerocaptures. D2 shuttles back and forth from LEO. What am I missing?

NASA has always talked about lunar exploration being a stepping-stone to Mars, and although that's not true in many ways, the one way it is true and relevant to SpaceX is that it's an opportunity to do *many* re-entry tests of Starship's heat shield at interplanetary velocities. Not with anyone onboard of course.

It seems to me that SpaceX would say using D2 as a lifeboat has about as much merit as designing a launch escape system for starship. Better to put more effort into making the main ship reliable.

That said, Earth EDL is the most intense challenge for starship's heat shield, so keep D2 as a proven Earth to LEO shuttle until crewed/P2P starship proves itself.

Since this CONOPS works, a CONOPS that uses two HLSs will also work without the need for a crewed refill. Just do your mission, but also send a second HLS as an OTV to that same LLO and refill it so it is waiting for the actual HLS when it ascends. crew transfers to the OTV and goes back to LEO. HLS and Depot make their way back to LEO more or less as you describe for Depot.

Disadvantage: needs more tanker flights and a second HLS.
Advantages:
   1) Removes crewed refill from the argument. This is more political than  technical.
   2) Provides for rescue in lots of failure scenarios.
   3) A variant can replace the SLS/Orion functionality even when the lander is Blue Moon Mk II (Artemis V).

[crandles57]

Since this CONOPS works, a CONOPS that uses two HLSs will also work without the need for a crewed refill. Just do your mission, but also send a second HLS as an OTV to that same LLO and refill it so it is waiting for the actual HLS when it ascends. crew transfers to the OTV and goes back to LEO. HLS and Depot make their way back to LEO more or less as you describe for Depot.

Disadvantage: needs more tanker flights and a second HLS.
Advantages:
   1) Removes crewed refill from the argument. This is more political than  technical.
   2) Provides for rescue in lots of failure scenarios.
   3) A variant can replace the SLS/Orion functionality even when the lander is Blue Moon Mk II (Artemis V).

How does the docking work here? If your HLS docks to you OTV in lunar orbit/NRHO then can dragon dock to each for LEO astronaut transfers?

Not sure if there is much if any merit and certainly a downside in eventually expending a dragon to the following:

Prepare HLS demo, HLS return vehicle, HLS starships.
Fully fuel HLS demo and HLS return vehicle.
Dock a dragon to HLS demo
HLS demo and HLS return vehicle TLI to NRHO (tests out dragon on nose through TLI)
Undock dragon
HLS demo doesn't need all its fuel so transfer some to HLS return vehicle to test out fuelling in NRHO.
Dock dragon to HLS return vehicle to test out this docking
HLS demo does lunar landing, lunar ascent and re-land tests
Fully fuel HLS
Launch astronauts to HLS in LEO dragon.
LEO dragon undocks and stays in LEO
TLI of HLS to lunar orbit, lunar landing & mission
HLS lunar ascent to NRHO.
Use lunar dragon to transfer to HLS return vehicle
TEI and propulsive burn to LEO
Dock with dragon in orbit for return to Earth

May need a 3rd dragon on standby in case of emergency return at a different time that cannot dock with dragon in LEO.

No refuelling in lunar orbit needed. Possible exception: If the already tested dragon astronaut transfer somehow fails, could do refuel HLS from HLS return vehicle.

Pro
At least the HLS and HLS return vehicle can be the same. (i.e. no male/female docking port differences needed)
Dragon left in NRHO might be useful for future astronaut and supply transfers?
Lunar dragon only used for astronaut transfers in NRHO which may need little adaptation, testing and verification? TLI test is only on an uncrewed HLS demo starship.
Gets lots of docking and refuelling in NRHO tests done.

Cons
Heavy on Dragon use, including eventually expending lunar dragon (unless it can somehow be refuelled)
2 HLS type ships needed

[DanClemmensen]

How does the docking work here? If your HLS docks to you OTV in lunar orbit/NRHO then can dragon dock to each for LEO astronaut transfers?

I have always assumed that HLS has an IDSS port that can act as either active or passive ("active/passive IDSS"), and SpaceX' latest document explicitly states that this is the case. Recall that the IDSS spec provides for this. the IDAs on ISS are passive-only. Dragon 2 is active-only. HLS is active/passive. the document states that the HLS port derives from the Dragon 2 IDSS by the addition of passive capability. There is also a picture of the port undergoing a mating test on the ground and I'm not sure but that picture seems to show both sets of latches.

HLS has had this requirement since NASA declared that Gateway might or might not be available for Artemis III. If available, HLS must be active. If not available, HLS must be passive to dock with Orion. There are some seriously kludgy alternatives, but active/passive is the simplest.

[jstrotha0975]

Could SpaceX modify a Crew Dragon pressure vessel to a lunar lander?

[Robotbeat]

Almost anything is possible. Not everything is beneficial. (To borrow a phrase)

[M.E.T.]

Why even do refuelling in lunar orbit if it is deemed risky with crew on board? Why not just put a small crew cabin on the fuel depot ship. Then HLS can just launch from the Moon and transfer the crew to the depot ship which already has all the fuel on board and the depot can take them back to LEO where Dragon awaits them.