SpaceX to NASA quote : simplified mission architecture : Technical discussion

[TheRadicalModerate]

First, I've already said the first missions will be refueled with fuel launched from LEO.  Yes we did go wildly off topic speculating about long term stuff, but no ISRU - probably not for the first decade.

If you insist on waiting for an EDLC-HLS (EDL-Capable HLS), why wouldn't you just refuel in HEEO before TLI?¹  Unlike surface refueling, you can actually get the tanker back to Earth that way, and it's considerably cheaper in terms of prop to LEO.  See attached for both cases.  Note that the surface refueling case doesn't close.


I'd like to drill down on why you think HLS as currently specified is so off-track for Mars.  The main differences from Pez Starship are:

1) No TPS.  But an insulating coating is substituted, which is needed for the depot--and is therefore 100% applicable to Mars refueling.

2) No header tanks and a nose dock.  This is indeed work that wouldn't be needed for Mars.  But the lack of header tanks is subtractive work--it doesn't cost much.  The nose dock is indeed new work, but it's presumably merely an enhancement to the existing D2 dock.

3) Landing thrusters.  These are almost certainly needed for Mars, for the same reasons they're needed for the Moon.  Furthermore, 50-100kN thrusters are likely needed for coarse attitude control and small-delta-v maneuvers.  Those are both applicable to Mars.

4) Landing legs.  The lunar ones don't need TPS and the Mars ones do, but that seems like something you'd plan for in the design.  I'd guess that the work is 90% common.

5) Crew module.  HLS with a crew of 4 probably only needs 120 crew-days of ECLSS and consumables, while Mars, for a crew of 10, needs something like 8000 crew-days.  That's obviously different.  But is there any universe where you go from 0 crew-days to 8000 crew-days without at least a couple of intermediate steps? 120 sounds like a pretty good first intermediate number.

6) Reusability.  The fact that an HLS starts out as expendable doesn't mean that SpaceX has to put aside reusability.  It's still going on with tankers and Pez Starships.  It's an expense not to reuse an HLS, but it's also an expense that NASA will happily pay SpaceX to incur.

7) Crew-rating launch and EDL.  Yup, those are things that you eventually need for Mars, and you don't need to HLS.  But once again, this is a subtractive requirement.  You'd still be doing all the same work on the crew module for launch/EDL crew-rating as you're doing for the crew module for HLS.  It is true that HLS requires figuring out lunar landing and ascent, which is different on Mars.  But your argument isn't that SpaceX should stop doing the Moon; instead, you're arguing that they should just do the Moon in a way that's completely congruent, right from the git-go, with how crewed Mars missions will happen.

IMO, Starship isn't going to be crew-certified for launch/EDL until at least 2030.  Crewed missions to Mars aren't going to happen before 2037.  Don't you think it makes sense to do something useful to human spaceflight in the interim?  90% of the DDT&E costs of HLS apply to Mars.  They should do that.

Second, the way I parse "faster return to the moon" is "faster than the real schedule", not "faster than the fake schedule we have been telling congress and fooling ourselves with".

Assumes facts not in evidence.  And it's usually not a great PR move to claim that you were talking about the secret schedule when it turns out that you knew you were three years behind the public one.  The SpaceX statement is in plain language.  It should be interpreted using the meaning of the plain language.

SpaceX might think that they're not going to make the schedule for the project as specified.  But they're almost certainly working toward it as specified.  And if they come up with a simplification/acceleration, it'll be something that fulfills the requirements, as specified.

___________
¹PS:  I'm a notorious hand-wringer on the ease of doing HEEO refuelings, because crewed rendezvous with high eccentricity, in a small number of orbits (to minimize Van Allen radiation exposure of the crew), isn't going to be a slam-dunk.  But it is a slam-dunk compared to transporting prop from Ship to Ship on the lunar surface.

[InterestedEngineer]

___________
¹PS:  I'm a notorious hand-wringer on the ease of doing HEEO refuelings, because crewed rendezvous with high eccentricity, in a small number of orbits (to minimize Van Allen radiation exposure of the crew), isn't going to be a slam-dunk.  But it is a slam-dunk compared to transporting prop from Ship to Ship on the lunar surface.

What happens if HSL has a fuel leak on the surface and doesn't have enough fuel to lift off and get back to earth?

[Lampyridae]

3) Landing thrusters.  These are almost certainly needed for Mars, for the same reasons they're needed for the Moon.  Furthermore, 50-100kN thrusters are likely needed for coarse attitude control and small-delta-v maneuvers.  Those are both applicable to Mars.

___________
¹PS:  I'm a notorious hand-wringer on the ease of doing HEEO refuelings, because crewed rendezvous with high eccentricity, in a small number of orbits (to minimize Van Allen radiation exposure of the crew), isn't going to be a slam-dunk.  But it is a slam-dunk compared to transporting prop from Ship to Ship on the lunar surface.

Do we know anything about these landing thrusters? Methalox or hypergolic? Because they are essentially descent/ascent engines for free. Detach the main tankage and point down (mostly) to reduce cosine losses, +/- LOR with another main tankage section/refuel/dock with D3 and you get a variety of options from crasher stage to direct Earth orbit return.

[TheRadicalModerate]

___________
¹PS:  I'm a notorious hand-wringer on the ease of doing HEEO refuelings, because crewed rendezvous with high eccentricity, in a small number of orbits (to minimize Van Allen radiation exposure of the crew), isn't going to be a slam-dunk.  But it is a slam-dunk compared to transporting prop from Ship to Ship on the lunar surface.

What happens if HSL has a fuel leak on the surface and doesn't have enough fuel to lift off and get back to earth?

What happens if HLS has a fuel leak in LLO?  Or TLI, for that matter?

If you want to put a spare HLS on the surface, as far away as possible, to increase the likelihood of it surviving the landing of the primary, you can do that, for some large expense. But you’ve been talking about landing prop for a refueling operation. That’s a giant bolus of new tech, and it still requires expending the tanker you land, which is the thing you wanted to avoid.

[TheRadicalModerate]

Do we know anything about these landing thrusters? Methalox or hypergolic? Because they are essentially descent/ascent engines for free. Detach the main tankage and point down (mostly) to reduce cosine losses, +/- LOR with another main tankage section/refuel/dock with D3 and you get a variety of options from crasher stage to direct Earth orbit return.

Short answer: no, we don't know anything.  Longer answer:

We can estimate the total thrust and prop mass that needs to be devoted to landing the last few hundred meters and ascending to a safe height to restart the engines on ascent.

At first blush, it looks like a temporary solution could be to use SuperDracos, which have about the right thrust for the requirements.  You'd need to load MMH/NTO, which is a pain, and you couldn't reuse the HLS, but that's not the plan for the first few missions.  However, two problems:

1) You need to restart them for ascent, so you'd have to fix the kludge with the burst-discs, which solved the backflow problem.

2) They're pressure-fed, and when you crank through the amount of prop you need, it's a surprisingly large number.  Pressure-fed systems' mass have a pretty steep knee in the curve as the prop volume goes up.

There are whole threads about Starship thrusters elsewhere--too lazy to look them up.  None of them got much traction, because nobody knows nuthin'.  I'd sorta convinced myself that the landing thrusters (which can do triple duty as thrusters for coarse attitude control and small-delta-v maneuvers), needed to be pump-based.  That sounds terrible, so I hope I'm wrong.

[Twark_Main]

1) You need to restart them for ascent, so you'd have to fix the kludge with the burst-discs, which solved the backflow problem.

The "backflow problem" was just operator error on the test stand, and since there were no restarts they just switched to burst discs out of an abundance of caution.

So the "fix" is just to switch back to the old design and put additional operational safeguards in place. There's no hardware fix, because it wasn't caused by a hardware failure.

[dchenevert]

The "backflow problem" was just operator error on the test stand, and since there were no restarts they just switched to burst discs out of an abundance of caution.

has "abundance of caution" become official tongue-in-cheek snark, w.r.t. space organizations that are failing to deliver, claiming "abundance of caution"? (asking for a legacy prime)

[Vultur]

Isn't SpaceX also still kind of hoping to avoid needing landing thrusters at all? At least, that's how I read the comment about studying Raptor plume/surface interaction in the "Moon and beyond" post. I think that's one possible "simplification", if it works.

[rsdavis9]

Isn't SpaceX also still kind of hoping to avoid needing landing thrusters at all? At least, that's how I read the comment about studying Raptor plume/surface interaction in the "Moon and beyond" post. I think that's one possible "simplification", if it works.

I read this as for mars landing thrusters are not needed because the soil is very different.
On the moon "I guess we will need them for at least the first landing" after that a landing pad would help.

I am not an expert on this. I am only guessing.

[TheRadicalModerate]

Isn't SpaceX also still kind of hoping to avoid needing landing thrusters at all? At least, that's how I read the comment about studying Raptor plume/surface interaction in the "Moon and beyond" post. I think that's one possible "simplification", if it works.

I read this as for mars landing thrusters are not needed because the soil is very different.
On the moon "I guess we will need them for at least the first landing" after that a landing pad would help.

I am not an expert on this. I am only guessing.

Your guesses mostly jibe with my guesses, although I think a landing pad (actually, I suspect a landing tower would be easier) is a ways off, and I suspect that thrusters will be needed for Mars.  However, it shouldn't be hard to double the number of thrusters around the waist.

Note that landing thrust is an area where Blue has a substantial advantage over SpaceX.  I'm pretty sure that 3 BE-7's won't excavate under the BM1.5 or BM2.

[TheRadicalModerate]

1) You need to restart them for ascent, so you'd have to fix the kludge with the burst-discs, which solved the backflow problem.

The "backflow problem" was just operator error on the test stand, and since there were no restarts they just switched to burst discs out of an abundance of caution.

So the "fix" is just to switch back to the old design and put additional operational safeguards in place. There's no hardware fix, because it wasn't caused by a hardware failure.

I thought it was a leaky check valve between the NTO tank and the He pressurant system.  That's not operator error.

They replaced the check valve with a burst disc, which the He pressurant bursts when the SDs need to run.  That keeps backflow out of system, but it makes it a one-shot.

I suspect they could replace the He pressurant valve with something that didn't slam open and then verify that the more gradual pressurization met whatever the reliability requirement was for the SDs, but that sounds like a fix to me--no scare quotes involved.

[StraumliBlight]

I thought it was a leaky check valve between the NTO tank and the He pressurant system.  That's not operator error.

They replaced the check valve with a burst disc, which the He pressurant bursts when the SDs need to run.  That keeps backflow out of system, but it makes it a one-shot.

I suspect they could replace the He pressurant valve with something that didn't slam open and then verify that the more gradual pressurization met whatever the reliability requirement was for the SDs, but that sounds like a fix to me--no scare quotes involved.

https://web.archive.org/web/20190731060740/https://www.spacex.com/news/2019/07/15/update-flight-abort-static-fire-anomaly-investigation

Initial data reviews indicated that the anomaly occurred approximately 100 milliseconds prior to ignition of Crew Dragon’s eight SuperDraco thrusters and during pressurization of the vehicle’s propulsion systems. Evidence shows that a leaking component allowed liquid oxidizer – nitrogen tetroxide (NTO) – to enter high-pressure helium tubes during ground processing. A slug of this NTO was driven through a helium check valve at high speed during rapid initialization of the launch escape system, resulting in structural failure within the check valve. The failure of the titanium component in a high-pressure NTO environment was sufficient to cause ignition of the check valve and led to an explosion.

In order to understand the exact scenario, and characterize the flammability of the check valve’s titanium internal components and NTO, as well as other material used within the system, the accident investigation team performed a series of tests at SpaceX’s rocket development facility in McGregor, Texas. Debris collected from the test site in Florida, which identified burning within the check valve, informed the tests in Texas. Additionally, the SuperDraco thrusters recovered from the test site remained intact, underscoring their reliability.

It is worth noting that the reaction between titanium and NTO at high pressure was not expected. Titanium has been used safely over many decades and on many spacecraft from all around the world. Even so, the static fire test and anomaly provided a wealth of data. Lessons learned from the test – and others in our comprehensive test campaign – will lead to further improvements in the safety and reliability of SpaceX’s flight vehicles.

Report seems to have been removed from the SpaceX website.

[Vultur]

Isn't SpaceX also still kind of hoping to avoid needing landing thrusters at all? At least, that's how I read the comment about studying Raptor plume/surface interaction in the "Moon and beyond" post. I think that's one possible "simplification", if it works.

I read this as for mars landing thrusters are not needed because the soil is very different.
On the moon "I guess we will need them for at least the first landing" after that a landing pad would help.

I am not an expert on this. I am only guessing.

I think that has been said at some point, but the recent Moon and Beyond" update refers to "Raptor lunar landing throttle test demonstrating a representative thrust profile that would allow Starship to land on the lunar surface".

That looks to me like SpaceX is still hoping to land on the Moon (not just Mars) with Raptors, and may be trying to demonstrate this at least partly with the intent to offer "no landing thrusters needed" to NASA as a potential simplification.

[TheRadicalModerate]

I think that has been said at some point, but the recent Moon and Beyond" update refers to "Raptor lunar landing throttle test demonstrating a representative thrust profile that would allow Starship to land on the lunar surface".

That looks to me like SpaceX is still hoping to land on the Moon (not just Mars) with Raptors, and may be trying to demonstrate this at least partly with the intent to offer "no landing thrusters needed" to NASA as a potential simplification.

It's absolutely a requirement to use Raptors to land on the Moon.  What's at issue is whether to use Raptors for the last hundred meters or so.  But there's still a throttle profile that's required.

Even if you're using thrusters, you need the Raptors to deliver the HLS to a specific height with a specific downward velocity.  If the velocity is too low, the thrusters have to use a lot of extra prop, and if they get close to running out, that's an abort.  If the velocity is too high, then the thrusters' thrust may not be adequate to do a safe hoverslam, and that's also an abort--or a crash.

[crandles57]

Even if you're using thrusters, you need the Raptors to deliver the HLS to a specific height with a specific downward velocity.  If the velocity is too low, the thrusters have to use a lot of extra prop, and if they get close to running out, that's an abort.

If the downward velocity is too low don't you turn down the throttle level, saving prop, until the downward velocity increases towards the ideal profile downward velocity for the height above surface? I guess that if you cannot turn the throttle level down much then it is going to take longer and the extra time (more prop) may well be more important than the throttle level (prop saving). This explanation seems to require that you are close to minimum throttle levels?

Another possible explanation might be if the downward velocity is too low then maybe you are longer before landing and that might involve drifting past the intended landing spot and more fuel is required finding and adjusting to a suitable landing location?

[TheRadicalModerate]

Even if you're using thrusters, you need the Raptors to deliver the HLS to a specific height with a specific downward velocity.  If the velocity is too low, the thrusters have to use a lot of extra prop, and if they get close to running out, that's an abort.

If the downward velocity is too low don't you turn down the throttle level, saving prop, until the downward velocity increases towards the ideal profile downward velocity for the height above surface? I guess that if you cannot turn the throttle level down much then it is going to take longer and the extra time (more prop) may well be more important than the throttle level (prop saving). This explanation seems to require that you are close to minimum throttle levels?

I think you've got it right:  You can throttle down if velocity is too low, but you still need the thruster burning, so there's some minimum amount of prop that must be used.

Note that prop minimization implies using the highest velocity and lowest altitude possible at Raptor cutoff / thruster start.  Then you increase altitude and/or decrease velocity until the mission safety people no longer have brown trousers.  But that should still yield solutions where the thrusters are close to their maximum throttle levels.

Another possible explanation might be if the downward velocity is too low then maybe you are longer before landing and that might involve drifting past the intended landing spot and more fuel is required finding and adjusting to a suitable landing location?

I'd think that you'd rely on Raptors to do any horizontal corrections to the landing spot, in any direction, before doing the switchover to the thrusters.

Note that I'm assuming that shutting down the Raptors without successfully starting the thrusters is a no-go.  If you can do that, then you could hover on Raptors at some fairly high altitude (a few hundred meters), translate until you find a good spot, null out any drift, kill the Raptors, fall for a while, then fire up the thrusters just in time for an optimal hoverslam.  But that has serious brown trouser problems.

[DanClemmensen]

Even if you're using thrusters, you need the Raptors to deliver the HLS to a specific height with a specific downward velocity.  If the velocity is too low, the thrusters have to use a lot of extra prop, and if they get close to running out, that's an abort.

If the downward velocity is too low don't you turn down the throttle level, saving prop, until the downward velocity increases towards the ideal profile downward velocity for the height above surface? I guess that if you cannot turn the throttle level down much then it is going to take longer and the extra time (more prop) may well be more important than the throttle level (prop saving). This explanation seems to require that you are close to minimum throttle levels?

I think you've got it right:  You can throttle down if velocity is too low, but you still need the thruster burning, so there's some minimum amount of prop that must be used.

Note that prop minimization implies using the highest velocity and lowest altitude possible at Raptor cutoff / thruster start.  Then you increase altitude and/or decrease velocity until the mission safety people no longer have brown trousers.  But that should still yield solutions where the thrusters are close to their maximum throttle levels.

Another possible explanation might be if the downward velocity is too low then maybe you are longer before landing and that might involve drifting past the intended landing spot and more fuel is required finding and adjusting to a suitable landing location?

I'd think that you'd rely on Raptors to do any horizontal corrections to the landing spot, in any direction, before doing the switchover to the thrusters.

Note that I'm assuming that shutting down the Raptors without successfully starting the thrusters is a no-go.  If you can do that, then you could hover on Raptors at some fairly high altitude (a few hundred meters), translate until you find a good spot, null out any drift, kill the Raptors, fall for a while, then fire up the thrusters just in time for an optimal hoverslam.  But that has serious brown trouser problems.

I think you want to use Raptors to "hoverslam" to a zero velocity at about 100 meters altitude at the instant the last raptor shuts off. Note that since you are in a lunar orbit, you are nearly horizontal at this point. Thrusters flip to vertical and you hover if needed to make the final centimeter-precision landing location.

[TheRadicalModerate]

I think you want to use Raptors to "hoverslam" to a zero velocity at about 100 meters altitude at the instant the last raptor shuts off.

It takes a lot of thruster propellant to do that, assuming you start the thrusters when you shut down the Raptors.  (As I said above, I suspect that mission safety rules will require this.)

If you have thrusters that could throttle down to 10% or so, the prop loss might not be too bad.  Otherwise, you want a pretty decent downward velocity (5-10m/s?) at the switchover.

The other variable in this is whether the thrusters are pumped or pressure-fed.  If they're pumped, they can probably use prop directly from the mains.  If they're pressure-fed, there have to COPVs and pressurant vessels, and they get heavy pretty quick.

Note that since you are in a lunar orbit, you are nearly horizontal at this point. Thrusters flip to vertical and you hover if needed to make the final centimeter-precision landing location.

Lunar descent is basically a reverse gravity turn.  100m up, the HLS should be within a couple of degrees of vertical, if not perfectly vertical.  Thrusters, especially waist-mounted ones, won't have the oomph to overcome the HLS's huge moment of inertia.

[Brigantine]

Lunar descent is basically a reverse gravity turn.

Yes, possibly

100m up, the HLS should be within a couple of degrees of vertical

No. Well, at least not as an inevitable feature of a reverse gravity turn - only as a design choice.

The near-ground portion of gravity turns look very different if they're optimized for no air resistance (no max-Q to consider) and very high thrust:weight (weight calculated in lunar gravity, so 6x less)

What's the relevant TWR here? 6 raptors (sea level engines at minimal thrust) ≈ 12 MN thrust vs 700 tons ≈ 1.2 MN weight, 10:1? So you can be pitched over 84⁰ from vertical and still support the full weight at full thrust. (descend at constant rate). It won't be done that way, but that's a relevant anchor to have.

IMO the moment of inertia and acceptable angular rates will be limiting factors in the design of a safe-yet-efficient descent profile - in lieu of aerodynamic considerations. The reverse gravity turn will be designed around the needs of the dramatic pitch-up-maneuver. (and then visual navigation requirements)

But this should all be the same as PoR, no?

[Twark_Main]

Lunar descent is basically a reverse gravity turn.

Yes, possibly

100m up, the HLS should be within a couple of degrees of vertical

No. Well, at least not as an inevitable feature of a reverse gravity turn - only as a design choice.

The near-ground portion of gravity turns look very different if they're optimized for no air resistance (no max-Q to consider) and very high thrust:weight (weight calculated in lunar gravity, so 6x less)

What's the relevant TWR here? 6 raptors (sea level engines at minimal thrust) ≈ 12 MN thrust vs 700 tons ≈ 1.2 MN weight, 10:1? So you can be pitched over 84⁰ from vertical and still support the full weight at full thrust. (descend at constant rate). It won't be done that way, but that's a relevant anchor to have.

...

Yes. Technically speaking, the most efficient lunar landing trajectory would be to

      1. Lower your perigee to the minimum acceptable height above the surface.

      2. Burn retrograde to circularize.

      3. Continue burning mostly retrograde, but tilting slightly (reaching a maximum of 6° cosine loss right before cutoff) to support your weight as you are now below orbital velocity.

      4. Proceed with your terminal landing sequence. Ideally this would be a rotation initiated by the engines right before cutoff, which reaches vertical right as the landing engines start for the hoverslam.

This maximizes Oberth because the burn is performed as low in the gravity well as possible.