Replacing SLS/Orion using Starship HLS and Crew Dragon

[TheRadicalModerate]

I'm gonna cross-post this over here to see if somebody has any opinions:

A question:  Do you think HLS Starship will be based on Starship v2 or v3?

Starship v2:
Pros:
- Design freezes earlier (probably desirable for crew-rating paperwork and human factors studies).
- Lower center of mass.
Cons:
- Assuming boiling prop at LEO departure (IMO a good assumption), it can't exceed 110t of inert (100t dry + 10t crew module) mass.  (This has a bunch of boiloff assumptions built into it, and it assumes that most big burns are handled wtih 3 RVacs + 1 RSL @ 50% throttle, for an average Isp of 367s.)
- At 10t crew module mass, there's not a lot of margin for the ECLSS to be able to handle longer missions, which would be handy if SpaceX plans to replace SLS/Orion with D2 + an orbital transfer version of HLS Starship.
- This assumes that cryo insulation/MMOD tiles¹ only have about a third of the TPS density, and only cover the tanks.

Starship v3:
- Pros:
- You've got big, sloppy margins for the crew module, ECLSS inside the crew modules, consumables, and payload.
- Boil-off margins are much bigger.
- Slightly higher Isp (369s?)
- You can slightly lower the CoM by removing payload barrel segments, but LCH4 sitting in the common dome doesn't move.
Cons:
- Higher center of mass on landing.
- Higher fueling costs. (More tankers to LEO.)
- Greater schedule risk.  (Added from original list.)

____________
¹See here for why I think we're now dealing with tiles instead of just Solar White paint.

One of the possible architectures that's been floated up-thread is to use a single LSS-HLS, fly it all the way from LEO to the lunar surface, ascend back to some kind of lunar orbit, refuel it, and then proceed back to LEO.  To use a v2 LSS for this requires a refueling in high LEO and another in either NRHO or LLO, after the ascent from the lunar surface.  That in turn requires a depot or tanker to shuttle the prop up to the high LEO. 

On the other hand, a v3 LSS could be refueled once in VLEO, and once in lunar orbit.  You don't have to shuttle prop up to high LEO.

[TheRadicalModerate]

Everybody seems hell-bent on finding a way to help NASA save face, even if it means constantly funding a program that is wastefully consuming enormous amounts of taxpayer money trying to reach a goal of landing on the moon while the goal posts are moving to the right faster than the calendar changes pages.

Way far off-topic.  There are other threads for this.

Actually not off-topic at all.
Topic Title: "Replacing SLS/Orion using Starship HLS and Crew Dragon".
This is exactly what I suggested doing:

1. Send Starship HLS into LEO to be refueled by tankers (already required by Artemis).
2. Send Crew Dragon into LEO to meet up with Starship HLS.
3. Crew transfers from Crew Dragon into Starship HLS and goes directly to the Lunar surface.
4. Execute Lunar surface mission.
5. Starship HLS departs Lunar surface and goes directly to LEO to meet the waiting Crew Dragon.
6. Crew transfers from Starship HLS into the Crew Dragon and returns to Earth surface.
7. Starship HLS remains in LEO to be refueled and reused for the next Lunar mission.

I think I replaced SLS/Orion with Starship HLS and Crew Dragon, just like the thread title.
What am I missing?

I don't object to the conops.  But I do object to the ranting about NASA.  It's fair to question whether the lunar staging orbit is NRHO or LLO if there are substantive differences in capabilities (and there are, especially if the LSS is going to be based on v2).  But let's just take it as given that SLS/Orion is out of the picture for purposes of this thread, and leave the political stuff alone.

You seem to be arguing two different conops without identifying which is which:

1) Use D2 as an Orion replacement, where it's what goes from LEO to lunar orbit and back to EDL.  This won't work without substantial modification of D2 and, per the OP, is off-topic.

2) Use D2 only as a shuttle to/from LEO, then use a single LSS-HLS to do LEO-LS-LEOpropulsive.  This will work and is on-topic, but it requires two refuelings:  one in VLEO (v3-based LSS) or high LEO (v2-based LSS), and a second refueling in lunar orbit, post-ascent.

Note that #2 still suffers from the RAAN mismatch problem for early aborts.

[sdsds]

Is the idea of using a single D2 per mission intended to reduce the amount of "Dragon time" consumed by each mission? Or simply to reduce the costs of launching, recovering and refurbishing a Dragon?

Dragon time on orbit is Dragon time wasted. The easily imaginable alternative is to launch on a D2, transfer the crew, and then recover the D2 immediately, getting it into the refurbish/re-launch process. Then launch a second D2 once the crew return time and orbital elements are certain.

I assert the marginal costs of refurbishing and re-launching a D2 are small compared with the overall cost of each mission, and during the mission the safest location for the crew re-entry D2 is on the ground, ready to fly. Even someone who doesn't believe safety should always be the first consideration will understand that from a political perspective it's more difficult to move forward a plan with a safety concern.

[TheRadicalModerate]

Orion aerobrakes.  Dragon2 Aerobrakes.  You can't go backwards on something that worked since the early 1960s that is that efficient. However you wish return from the moon you have to involve aerobraking.

If you use that terminology, things will get confusing later on.  I suggest the following:

1) EDL:  Just what the acronym says:  Entry, descent, and landing.  Speed scrubbed off:  >7700m/s from LEO, >11,000m/s from TEI.

2) Aerocapture:  Single-pass through the atmosphere and back into space, with enough speed scrubbed off to be in something resembling the target orbit.  Speed scrubbed off (from TEI):  ~3100m/s.  Then you need ~60m/s of propulsive delta-v to raise the perigee.¹

3) Aerobraking:  Multiple passes through the atmosphere and back into space, with speed scrubbed off very gradually, until the apogee is right.  Speed scrubbed off:  could be as little as 25m/s per pass.  Then you do an apogee burn to raise the perigee when you get to the right apogee.  (At 25m/s per pass, it takes about 2 months to get a spacecraft from TEI into its parking orbit.)

For our purposes:

EDL-Capable spacecraft:  Orion, D2, EDLC (EDL-capable) Starship

Aerocapture-Capable spacecraft:  EDLC Starship (Orion and D2 have lost there service modules, and aren't viable in space any more.)

Aerobraking-Capable spacecraft:  EDLC Starship (but it's silly), maybe a depot, maybe an LSS-OTV.  LSS-HLS can't do it unless there's a solution to the solar panels on the nose.

___________
¹Note that, for Mars transits, you must at least aerocapture if you're not going to do a pure propulsive insert, because you have to drop below escape speed.  Then you can do additional aerobraking passes to shape the orbit if you'd like.  But if you have the TPS for aerocapture, you probably want to hit approximately the right apoapse during the capture.

[InterestedEngineer]

Orion aerobrakes.  Dragon2 Aerobrakes.  You can't go backwards on something that worked since the early 1960s that is that efficient. However you wish return from the moon you have to involve aerobraking.

If you use that terminology, things will get confusing later on.  I suggest the following:

1) EDL:  Just what the acronym says:  Entry, descent, and landing.  Speed scrubbed off:  >7700m/s from LEO, >11,000m/s from TEI.

2) Aerocapture:  Single-pass through the atmosphere and back into space, with enough speed scrubbed off to be in something resembling the target orbit.  Speed scrubbed off (from TEI):  ~3100m/s.  Then you need ~60m/s of propulsive delta-v to raise the perigee.¹

3) Aerobraking:  Multiple passes through the atmosphere and back into space, with speed scrubbed off very gradually, until the apogee is right.  Speed scrubbed off:  could be as little as 25m/s per pass.  Then you do an apogee burn to raise the perigee when you get to the right apogee.  (At 25m/s per pass, it takes about 2 months to get a spacecraft from TEI into its parking orbit.)

Those definitions are merely differences in method and deltaV.  In fact they are so specific they don't even cover the "slow down multiple passes and then EDL" case.   

What's the generic term for "use atmosphere to slow down a spacecraft"  AIR + BRAKE = aereobrake.  Like Humpty Dumpty, I'm just gonna use that term, because it follows the etymology of the words, not some head case in NASA's "muh 10,000 acronyms" opinion.

If you have a better term for the generic "use atmosphere to slow down a spacecraft" I'll be happy to use it, but I can't find it in the official literature.

EDIT:: Also it needs a verb form or the phrasing is going to get very awkward.  Oh, say like this:

https://www.dictionary.com/browse/aerobrake

[TheRadicalModerate]

I assert the marginal costs of refurbishing and re-launching a D2 are small compared with the overall cost of each mission, and during the mission the safest location for the crew re-entry D2 is on the ground, ready to fly. Even someone who doesn't believe safety should always be the first consideration will understand that from a political perspective it's more difficult to move forward a plan with a safety concern.

The costs aren't that small.  SpaceX charges NASA $300M for D2 missions to ISS.  I'd expect a D2 mission as a taxi to an LSS to be at least that expensive, and maybe more, due to what you said:  D2 time on orbit is D2 time wasted.

I don't think it'll be possible to land, refurbish, and re-launch a single D2 while the LSS is off at the Moon, so you need to reserve two, which complicates scheduling quite a bit (and therefore likely adds cost).

That said, even an extra $500M is mouse nuts compared to an SLS/Orion mission.

[InterestedEngineer]

I assert the marginal costs of refurbishing and re-launching a D2 are small compared with the overall cost of each mission, and during the mission the safest location for the crew re-entry D2 is on the ground, ready to fly. Even someone who doesn't believe safety should always be the first consideration will understand that from a political perspective it's more difficult to move forward a plan with a safety concern.

The costs aren't that small.  SpaceX charges NASA $300M for D2 missions to ISS.  I'd expect a D2 mission as a taxi to an LSS to be at least that expensive, and maybe more, due to what you said:  D2 time on orbit is D2 time wasted.

I don't think it'll be possible to land, refurbish, and re-launch a single D2 while the LSS is off at the Moon, so you need to reserve two, which complicates scheduling quite a bit (and therefore likely adds cost).

That said, even an extra $500M is mouse nuts compared to an SLS/Orion mission.

$500M is 10 non-2nd stage-reused launches by Starship, and likely by 2028 50 fully reused launches.

It'll be cheaper to test to failure Starship EDL than to put two D2s on this mission, as long as the failure count is < 10 (failure meaning the Starship can't be reused, not that it crashed and burned).

It'll also be  cheaper to launch a few extra full depots to LLO (and heck a spare HLS)  to refuel for the return trip and skip any aerobraking¹ to LEO, if one wants to be that paranoid about heat shields in 2028.

¹https://www.dictionary.com/browse/aerobrake

[DanClemmensen]

I assert the marginal costs of refurbishing and re-launching a D2 are small compared with the overall cost of each mission, and during the mission the safest location for the crew re-entry D2 is on the ground, ready to fly. Even someone who doesn't believe safety should always be the first consideration will understand that from a political perspective it's more difficult to move forward a plan with a safety concern.

The costs aren't that small.  SpaceX charges NASA $300M for D2 missions to ISS.  I'd expect a D2 mission as a taxi to an LSS to be at least that expensive, and maybe more, due to what you said:  D2 time on orbit is D2 time wasted.

I don't think it'll be possible to land, refurbish, and re-launch a single D2 while the LSS is off at the Moon, so you need to reserve two, which complicates scheduling quite a bit (and therefore likely adds cost).

That said, even an extra $500M is mouse nuts compared to an SLS/Orion mission.

When using two D2, the first D2 launch is a like a normal commercial free-flying D2 mission. These are cheaper than a six-month ISS mission, I think. The second D2 mission is more of a problem, because it must be on standby for the entire time of the HLS mission, and keeping it on standby means that it is not available for other missions. If the HLS crew declares an emergency, the Dragon must fly. I think you have about two days from the time the emergency is declared until the D2 should launch. This would be for a medical emergency where a crew member must return to Earth ASAP.

All of this takes a toll on the D2 fleet lifetime and the ability of D2 to support CLDs, and imposes a lower bound on the HLS mission cadence. Fortunately, SpaceX should be getting crewed EDL Starship certified before the D2 scheduling situation becomes critical.

[Twark_Main]

Crew Dragon might be able to return to Earth from LLO, but this is an untested capability, so it is off the table for this thread.

I respectfully disagree. Mr. Musk has repeatedly indicated in the past that the PICA-X heatshield is fully capable of interplanetary reentry velocity (25,000 mph) into earth's atmosphere. There simply hasn't been a need to demonstrate that; and that's all it needs.

A high-speed heatshield isn't the only thing needed.  You also need:

1) Upgraded ECLSS.  Instead of about 20 crew-days, it'll need at least 40.
2) Extended delta-v (and thrust) to allow a D2 to do the TEI burn.
3) Increased consumables space.
4) A solar storm shelter, or a viable solar storm configuration.
5) Electronics hardening, or at least testing in the deep space environment.

It's a lot of work to do it like the Apollo architecture.  Not impossible, but quite a distance from the current hardware.  It's hardly "as-is".

On the other hand, if you do the trick where you carry the D2 on the nose of the LSS-OTV and only use it to carry the crew to EDL at the last moment, you can eliminate items 1-4 above.  You'd still need an uncrewed test, but if, as you stated, the heatshield is good to go, then there's not much development to be done.

Seems like "increased consumables" and "solar storm shelter" are two problems that solve each-other.

I disagree that they need "increased consumables space."  The existing pressurized envelope has plenty of room for the incremental consumables.

[MickQ]

I assert the marginal costs of refurbishing and re-launching a D2 are small compared with the overall cost of each mission, and during the mission the safest location for the crew re-entry D2 is on the ground, ready to fly. Even someone who doesn't believe safety should always be the first consideration will understand that from a political perspective it's more difficult to move forward a plan with a safety concern.

The costs aren't that small.  SpaceX charges NASA $300M for D2 missions to ISS.  I'd expect a D2 mission as a taxi to an LSS to be at least that expensive, and maybe more, due to what you said:  D2 time on orbit is D2 time wasted.

I don't think it'll be possible to land, refurbish, and re-launch a single D2 while the LSS is off at the Moon, so you need to reserve two, which complicates scheduling quite a bit (and therefore likely adds cost).

That said, even an extra $500M is mouse nuts compared to an SLS/Orion mission.

When using two D2, the first D2 launch is a like a normal commercial free-flying D2 mission. These are cheaper than a six-month ISS mission, I think. The second D2 mission is more of a problem, because it must be on standby for the entire time of the HLS mission, and keeping it on standby means that it is not available for other missions. If the HLS crew declares an emergency, the Dragon must fly. I think you have about two days from the time the emergency is declared until the D2 should launch. This would be for a medical emergency where a crew member must return to Earth ASAP.

All of this takes a toll on the D2 fleet lifetime and the ability of D2 to support CLDs, and imposes a lower bound on the HLS mission cadence. Fortunately, SpaceX should be getting crewed EDL Starship certified before the D2 scheduling situation becomes critical.

Would it be economical to build a new D2 to have specifically on standby during lunar missions ??

[DanClemmensen]

I assert the marginal costs of refurbishing and re-launching a D2 are small compared with the overall cost of each mission, and during the mission the safest location for the crew re-entry D2 is on the ground, ready to fly. Even someone who doesn't believe safety should always be the first consideration will understand that from a political perspective it's more difficult to move forward a plan with a safety concern.

The costs aren't that small.  SpaceX charges NASA $300M for D2 missions to ISS.  I'd expect a D2 mission as a taxi to an LSS to be at least that expensive, and maybe more, due to what you said:  D2 time on orbit is D2 time wasted.

I don't think it'll be possible to land, refurbish, and re-launch a single D2 while the LSS is off at the Moon, so you need to reserve two, which complicates scheduling quite a bit (and therefore likely adds cost).

That said, even an extra $500M is mouse nuts compared to an SLS/Orion mission.

When using two D2, the first D2 launch is a like a normal commercial free-flying D2 mission. These are cheaper than a six-month ISS mission, I think. The second D2 mission is more of a problem, because it must be on standby for the entire time of the HLS mission, and keeping it on standby means that it is not available for other missions. If the HLS crew declares an emergency, the Dragon must fly. I think you have about two days from the time the emergency is declared until the D2 should launch. This would be for a medical emergency where a crew member must return to Earth ASAP.

All of this takes a toll on the D2 fleet lifetime and the ability of D2 to support CLDs, and imposes a lower bound on the HLS mission cadence. Fortunately, SpaceX should be getting crewed EDL Starship certified before the D2 scheduling situation becomes critical.

Would it be economical to build a new D2 to have specifically on standby during lunar missions ??

I don't know how expensive a new Crew Dragon is. I do know it's taken several years from the first public discussion of the fifth one until its actual availability. My impression (no source) is that they were trying to figure out if it would needed based on if/when Starliner would enter service. Since SpaveX plans to get Crewed EDL Starship operational before 2030, there is not much time for the fifth capsule to pay for itself.

[clongton]

Everybody seems hell-bent on finding a way to help NASA save face, even if it means constantly funding a program that is wastefully consuming enormous amounts of taxpayer money trying to reach a goal of landing on the moon while the goal posts are moving to the right faster than the calendar changes pages.

Way far off-topic.  There are other threads for this.

Actually not off-topic at all.
Topic Title: "Replacing SLS/Orion using Starship HLS and Crew Dragon".
This is exactly what I suggested doing:

1. Send Starship HLS into LEO to be refueled by tankers (already required by Artemis).
2. Send Crew Dragon into LEO to meet up with Starship HLS.
3. Crew transfers from Crew Dragon into Starship HLS and goes directly to the Lunar surface.
4. Execute Lunar surface mission.
5. Starship HLS departs Lunar surface and goes directly to LEO to meet the waiting Crew Dragon.
6. Crew transfers from Starship HLS into the Crew Dragon and returns to Earth surface.
7. Starship HLS remains in LEO to be refueled and reused for the next Lunar mission.

I think I replaced SLS/Orion with Starship HLS and Crew Dragon, just like the thread title.
What am I missing?

I don't object to the conops.  But I do object to the ranting about NASA.  It's fair to question whether the lunar staging orbit is NRHO or LLO if there are substantive differences in capabilities (and there are, especially if the LSS is going to be based on v2).  But let's just take it as given that SLS/Orion is out of the picture for purposes of this thread, and leave the political stuff alone.

Emphasis mine.
I understand what you're saying, I really, really do. To be clear, I have made it crystal clear in many different threads, many, many times, that my "ranting" about NASA is in the context of being a slave to Congress because of its funding profile. By law, NASA cannot do a single, solitary thing unless Congress gives it the money to to do it with. Today's NASA is totally and completely different from the NASA I grew up admiring and loving. Today's NASA is a congressional slave, run by people that must kowtow to its congressional masters, most of whom don't give a tinker's damn about what NASA is supposed to be doing, so long as congressional money (taxpayer dollars) continues to flow to their personal congressional districts. All NASA is to them is a vehicle to take our taxpayer dollars and send it to their own congressional districts, which then eventually finds its way back to them in campaign contributions to help fund election campaigns to keep them in power in congress. It's a giant, massive scam. I grew up admiring NASA like it was a god of some kind; until I actually had to deal with NASA myself. The truth is that NASA is a slave to people who really don't give a sh1t about you, me or what we want our civilian space agency to be doing. NASA is more of a victim than anything else, and I feel sorry for everyone who eventually has to see the agency for what it actually is, instead of the persona that it works SO very hard to present to the public. It's a very traumatic experience to see clearly what has become of the agency in lieu of what one has always believed it is.
And regrettably, there is no way to fix it.

But that is a topic for a different thread. I only outlined it here to address your valid thought you expressed in the hopes that you would understand where I'm coming from. So let's not take this any further here. It's already off topic and I want that to end here, and get back on topic.

[TheRadicalModerate]

Maybe it's time to update the taxonomy of things you can do with a D2 and one or more LSSes.  There are three basic architectures, each with a lot of conops variants.  First, the architectures, in their most favorable forms:¹

1) The Two-LSS Approach
D2 does launch-LEO-RPOD-loiterWeeks-RPOD-EDL
LSS-OTV does LEO-loiterMonths-refuel-RPOD-lunarOrbit-RPOD-loiterDays-RPOD-LEOpropulsive-RPOD-idle
LSS-HLS does lunarOrbit-loiterMonths-refuel-loiterMonths-RPOD-LS-surfaceOps-lunarOrbit-RPOD-loiterMonths

2) The One-LSS Approach
D2 does launch-LEO-RPOD-loiterWeeks-RPOD-EDL
LSS-HLS does LEO-loiterMonths-refuel-RPOD-LS-surfaceOps-lunarOrbit-refuel-LEOpropulsive-RPOD-loiterMonths

3) The One-LSS, D2 Direct EDL Approach
D2 does launch-LEO-RPOD-remainDocked-lunarOrbit-undock-loiterDays-RPOD-TEI-undockNearEntryInterface-directEDL
LSS-HLS does LEO-loiterMonths-refuel-RPOD-carryD2-lunarOrbit-undock(D2)-LS-surfaceOps-lunarOrbit-RPOD(D2)-TEI-undock(D2)-LEOaerobrake-circularize-loiterMonths


These are the three basic architectures.  However, lots of fiddly details can morph some of their conops into less favorable configurations.  So there are some trades to consider:

How do we haul prop to high orbit?
There are a number of options here:
a) Use a depot to haul it, and limit yourself to leaving enough prop for the depot to propulsively return to VLEO.
b) Use an expendable depot.
c) Use a tanker as a means of shuttling prop from a low to a high depot, then do direct EDL after transfer.

The costs of each of these aren't immediately obvious.  For purposes of analysis, I'm going to assume that reusable depots shuttle to HEEO and return propulsively, but tankers may shuttle to NRHO or LLO, leaving prop in a pre-positioned depot.  Also, I'm only going to consider cases where one shuttling flight occurs per mission.

Boil-off for tankers doing BLTs is an open question.  They should be able to point their noses at the sun and therefore have a fairly benign envrionment.

Do we need NRHO access to both drop off and retrieve Gateway Crew?  Do we need NRHO access at all?
For now, I'm gonna assume the answer is "yes" to both, except in a One-LSS Direct EDL cases.  If we're using v2 only, LEO-NRHO-LS-NRHO doesn't work at all, but LEO-LLO-LS-LLO does.  If we're using v3, we can visit NRHO only once, either before or after lunar descent, without a crewed refueling.  Both cases covered below.

A thought:  To the extent that we've considered the Gateway (other than getting rid of it), we've always assumed that crews were dropped off pre-descent and picked up post-ascent.  With the LSS-OTV, however, we don't have to do this; Gateway crews could be managed with missions separate from lunar surface missions.  Maybe they're piggybacked on resupply missions?

Are crewed refuelings OK?
If they're not OK, the only things that work are the Two-LSS Approach and the One-LSS with D2 Direct EDL, using a v3 propulsion section and only visiting NRHO once.²

Pre-descent or post-ascent refuelings in lunar orbit--or both?
If you're able and willing to do a crewed HEEO refueling, you never have to worry about pre-descent refuelings in lunar orbit.  That's good, because they're horrible in terms of prop consumption.  But post-ascent refuelings have the unfortunate problem that, if the refueling fails, the crew is stranded.  Note that stranded at the Gateway is better than plain ol' stranded.

There's also a weird but useful situation where you can send a single depot to NRHO, give the LSS-HLS a splash of prop pre-descent (just enough to get it down and back to NRHO), then do a post-ascent refueling that's large enough to get you back to LEO.  This is better than a full pre-descent refueling, because the LSS doesn't have to haul the LEO return prop down into the gravity well and back.

v2 or v3 Propulsion Sections
v2 works fine for the Two-LSS Approach, but the other two architectures really need v3.  For the approach where the D2 stays in LEO, it needs to be able to reach the LSS-HLS's refueling orbit if crewed refueling is to be avoided.  That limits it to 200x1400 eccentric orbit, if we use Polaris Dawn as an indicator of what's safely reachable. 

v3 LSS dry mass assumes that 3 payload barrel ring segments get removed, to reduce mass, increase tilt tolerance by a small amount, and to ease crew access to the garage.

I can't get either of the One-LSS Approaches to work within the D2's apogee at all, and the One-LSS with D2 Direct EDL only works for v3.  It is possible to keep the D2 in VLEO if we're willing to do three refuelings, one uncrewed in VLEO, one crewed in HEEO or NRHO, and another crewed post-ascent one in NRHO or LLO.

Is the RAAN problem a big deal?
If you don't know what this is, peruse the last few pages of this thread.  We know the RAAN problem can be solved by using a second D2 if necessary, but it's kludgy and moderately expensive.  If we wish to avoid the RAAN problem in all cases, the One-LSS Approach with D2 Direct EDL is the way to go.

Is crewed RPOD minimization a big deal?
The Two-LSS Approach requires four RPODs: two in LEO with the D2 and the LSS-OTV, and two in lunar orbit between the LSS-OTV and LSS-HLS.  The One-LSS approaches require anywhere from 4 to 6, with many of them being crewed refuelings.  With a v3 LSS, One-LSS with D2 Direct EDL and no stop in NRHO only needs 3 RPODs.  And the SLS/Orion cases are the best of all.

Roll-up of things that sorta work is attached, along with the Orion reference cases.  The hideous spreadsheet from which they're derived is here.  I'm sure there are a few mistakes; hope they're not too significant.

__________
¹Assumptions:
Crew Module mass = 15t
v2 LSS: Dry mass = 105t, Prop = 1400t (prop will be boiling and less dense!)
v3 LSS: Dry mass = 131t, Prop = 2146t
v2 Depot: Dry mass = 100t, Prop = 1788t (domes moved forward by 4 ring segments)
Average Isp = 369s, which may be optimistic

²If the D2 can reach a slightly higher energy than Polaris Dawn, it can do an RPOD with an LSS-HLS that's been topped off at an energy equivalent to either 840km circular or 200x1540 eccentric, and then reach NRHO.  That might be doable.  Note again that going straight through LLO has lots of margin.

[Greg Hullender]

a) Use a depot to haul it, and limit yourself to leaving enough prop for the depot to propulsively return to VLEO.

I'll bet you could aerocapture the depot, if you wanted to. It might take a dozen orbits (so two dozen transits of the Van Allen belts) but maybe that's okay.

[steveleach]

If something is risky and/or expensive, there are two options for dealing with it. Either do it as little as possible, and be very very careful, or do it as much as possible and get very very good at it.

I would not be at all surprised if SpaceX go with the latter for launching fuel to high energy orbits, and crewed refuelling / RPODs.

[DanClemmensen]

If something is risky and/or expensive, there are two options for dealing with it. Either do it as little as possible, and be very very careful, or do it as much as possible and get very very good at it.

I would not be at all surprised if SpaceX go with the latter for launching fuel to high energy orbits, and crewed refuelling / RPODs.

They might also initially use more conservative mission plans, using more fuel but fewer crewed refuelling ops, and/or using plans that allow for aborts or rescues if a fuelling fails non-catastrophically. As confidence grows with experience, shift to less conservative plans, perhaps with more complicated rescue options.

[steveleach]

If something is risky and/or expensive, there are two options for dealing with it. Either do it as little as possible, and be very very careful, or do it as much as possible and get very very good at it.

I would not be at all surprised if SpaceX go with the latter for launching fuel to high energy orbits, and crewed refuelling / RPODs.

They might also initially use more conservative mission plans, using more fuel but fewer crewed refuelling ops, and/or using plans that allow for aborts or rescues if a fuelling fails non-catastrophically. As confidence grows with experience, shift to less conservative plans, perhaps with more complicated rescue options.

Yeah,  I don't really expect them to do anything for the first time with crew aboard, unless there's absolutely no choice.

You've also got some interesting new contingency options for small crews on a vehicle as large as Starship, and that is expected to launch as often as Starship.

Including 18 months of supplies for a 4 week mission, for example, and then start designing & practicing rescue missions if a problem arises, knowing you've got a load of time to implement it.

Also, crewed Starship EDL might not be considered reliable enough for normal use, but you can bet it would be on the table if it was the best way to bring a stranded crew back, and if uncrewed EDL was at least reasonably reliable.

[TheRadicalModerate]

I'll bet you could aerocapture the depot, if you wanted to. It might take a dozen orbits (so two dozen transits of the Van Allen belts) but maybe that's okay.

Update: I posted the aerobraking idea over on the "Starship On-orbit refueling" thread.  Probably better to discuss over there.

Good point, and I'm already assuming that the LSS-HLS can be aerobraked for the One-LSS with D2 Direct EDL cases.  I should've tumbled to that idea.

With both, the unaddressed requirement is getting the solar panels to survive a set of even gentle aerobraking passes (<25m/s).  At the very least, letting them rip through ionized oxygen is going to significantly degrade their lifetime, but it's more likely they're gonna get destroyed if they're just stuck onto the nose somewhere.  That could be fixed by making them deployable/stowable from/to some kind of chine or doghouse, from the vestigial payload bay, or, if you're willing to depopulate a v3 back to 3 RVacs, from the skirt.  But for aerobraking, they probably need to be deployable tens to hundreds of times.  That's a bit more of a mechanical engineering task.

They might also initially use more conservative mission plans, using more fuel but fewer crewed refuelling ops, and/or using plans that allow for aborts or rescues if a fuelling fails non-catastrophically. As confidence grows with experience, shift to less conservative plans, perhaps with more complicated rescue options.

That's not always possible.

The primary limit for D2 + x conops is how high you can get the D2 safely.  If you can't do the rest of the conops on the completely refueled LSS at that energy, then you'll need to do a refueling with the crew on board, no matter how many tankers you send.

I thought this was going to be the principal advantage of using the 1-LSS with D2 direct EDL conops, but if you need to visit NRHO twice, this is exactly what you run afoul of:  you need the D2 to be able to reach a 200 x 1540km orbit, which is really close to what Polaris Dawn was able to do, but not quite.

I don't know if that apogee height was limited because of delta-v or radiation.  If it's the former, this conops will work with an expendable F9 booster.  I'd think that was cheaper than reserving a second D2 to deal with the RAAN problem, but it might be close.

[Twark_Main]

I'll bet you could aerocapture the depot, if you wanted to. It might take a dozen orbits (so two dozen transits of the Van Allen belts) but maybe that's okay.

Good point, and I'm already assuming that the LSS-HLS can be aerobraked for the One-LSS with D2 Direct EDL cases.  I should've tumbled to that idea.

With both, the unaddressed requirement is getting the solar panels to survive a set of even gentle aerobraking passes (<25m/s).  At the very least, letting them rip through ionized oxygen is going to significantly degrade their lifetime, but it's more likely they're gonna get destroyed if they're just stuck onto the nose somewhere.  That could be fixed by making them deployable/stowable from/to some kind of chine or doghouse, from the vestigial payload bay, or, if you're willing to depopulate a v3 back to 3 RVacs, from the skirt.  But for aerobraking, they probably need to be deployable tens to hundreds of times.  That's a bit more of a mechanical engineering task.

For Mars Global Surveyor they didn't retract the solar arrays during aerobraking. And standard LEO solar arrays are exposed to lots of atomic oxygen over their entire lifespan. Between the two, I expect it won't actually be an issue.

[meekGee]

I'll bet you could aerocapture the depot, if you wanted to. It might take a dozen orbits (so two dozen transits of the Van Allen belts) but maybe that's okay.

Good point, and I'm already assuming that the LSS-HLS can be aerobraked for the One-LSS with D2 Direct EDL cases.  I should've tumbled to that idea.

With both, the unaddressed requirement is getting the solar panels to survive a set of even gentle aerobraking passes (&lt;25m/s).  At the very least, letting them rip through ionized oxygen is going to significantly degrade their lifetime, but it's more likely they're gonna get destroyed if they're just stuck onto the nose somewhere.  That could be fixed by making them deployable/stowable from/to some kind of chine or doghouse, from the vestigial payload bay, or, if you're willing to depopulate a v3 back to 3 RVacs, from the skirt.  But for aerobraking, they probably need to be deployable tens to hundreds of times.  That's a bit more of a mechanical engineering task.

For Mars Global Surveyor they didn't retract the solar arrays during aerobraking. And standard LEO solar arrays are exposed to lots of atomic oxygen over their entire lifespan. Between the two, I expect it won't actually be an issue.

IIRC the MGS aerobraking was for orbit adjustment, wasn't it?  Not the actual capture?  They did many passes, shaving off just a bit at a time.

Aerocapture is at least one OOM more aggressive?