Starship Artemis Contract & Lunar Starship

[DanClemmensen]

A low-pressure gas-on-gas preburner to pump liquids into the thruster's main combustion chamber?

A low-pressure gas-on-gas burner to gassify LOX and methane?

[TheRadicalModerate]

A low-pressure gas-on-gas preburner to pump liquids into the thruster's main combustion chamber?

I think that's called a turbopump, although that's usually gas/liquid, not gas/gas.

The big, big advantage to pump-fed engines is they can be fed right from the mains, or at least a funny sump at the bottom of the mains, which you can think of as a propellant management device.

[Vultur]

Both.  (At the end of the day, there's not a huge difference.)

I don't see how there can not be a huge difference. One is technical, one is ultimately people/paperwork/bureaucratic.

They absolutely need Raptor/surface interaction tests, because they need to know where they have to shut off the Raptors, which controls how much prop the thrusters need.  But that could be more than 100m above the surface.

Sure.  But given that they've done that testing -- we don't know what it showed... But SpaceX does.

If SpaceX is convinced from that data that landing all the way with Raptors will work, I don't at all think that automatically convinces NASA.

But given that last fall NASA asked for options to accelerate/streamline lunar landing, I do think it's entirely possible that SpaceX offered landing on just Raptors as a streamlining. Depending on exactly when those tests were done, they may have presented that data to NASA then too.

(and it's possible that NASA didn't say no, even if they haven't officially said yes yet either, and that's why we aren't seeing landing thruster tests...)

Remember:  5.2GW of jet power for one Raptor, and a hoverslam on a rough surface with bad PNT.  "No" seems like a no-brainer to me.

I agree about the power, but not necessarily the other two.

They may be able to choose a surface that's not very rough, they'll have way better imagery than Apollo did. (Unless you just mean an unprepared surface, in which  case I do agree.)

I don't see why PNT needs to be that critical; if the sensors are only "blind" to the surface for the last little bit, and the vehicle knows its thrust (accelerometers), how will position errors build up in that small time to be beyond the tolerance of the legs (which could be significant, they're likely to be large legs)?

[MickQ]

I see one way to find out for sure.

Land first with thrusters, drop a rover or two with cameras to watch the ship take off and then land again on the raptors.

[daedalus1]

Isn't one Raptor far too powerful to land on lunar surface. More than 200 tonnes of thrust verses 1/6th weight of Starship?

[SpaceLizard]

Isn't Raptor throttleable down to something like 45% thrust?

[daedalus1]

Isn't Raptor throttleable down to something like 45% thrust?

That would still be too much.

[SpaceLizard]

Isn't Raptor throttleable down to something like 45% thrust?

That would still be too much.

According to Wikipedia Raptor 3 has 280 tons of thrust anyway so: 280 x 45% = 126 tons of thrust. Guess they'll "just" have to do a hoverslam, good thing they already have some practice with Falcon 9...

[daedalus1]

Isn't Raptor throttleable down to something like 45% thrust?

That would still be too much.

According to Wikipedia Raptor 3 has 280 tons of thrust anyway so: 280 x 45% = 126 tons of thrust. Guess they'll "just" have to do a hoverslam, good thing they already have some practice with Falcon 9...

Assuming a mass 200 tonnes that equates to a weight of about 35 tonnes. So far too much thrust.

[StraumliBlight]

According to Wikipedia Raptor 3 has 280 tons of thrust anyway so: 280 x 45% = 126 tons of thrust. Guess they'll "just" have to do a hoverslam, good thing they already have some practice with Falcon 9...

(160 tons (Starship mass) + 100 tons (payload mass) + 500 tons (fuel)) * 0.165 (lunar gravity) = 126 tf.

[daedalus1]

According to Wikipedia Raptor 3 has 280 tons of thrust anyway so: 280 x 45% = 126 tons of thrust. Guess they'll "just" have to do a hoverslam, good thing they already have some practice with Falcon 9...

(160 tons (Starship mass) + 100 tons (payload mass) + 500 tons (fuel)) * 0.165 (lunar gravity) = 126 tf.

Ok, I had no idea of the mass of fuel for return to orbit. But I  was calculating for a one way trip to the surface for use as habitat, so no fuel except for the final landing thrust.

[DanClemmensen]

According to Wikipedia Raptor 3 has 280 tons of thrust anyway so: 280 x 45% = 126 tons of thrust. Guess they'll "just" have to do a hoverslam, good thing they already have some practice with Falcon 9...

(160 tons (Starship mass) + 100 tons (payload mass) + 500 tons (fuel)) * 0.165 (lunar gravity) = 126 tf.

SpaceX can just tell NASA that the payload mass is required to be 200 tonne. 

It's a bit awkward to land HLS on a single Raptor, since none of the six Raptors are on the Ship's vertical axis. This means HLS would land at a tilt, and things will get interesting during the last meter of the landing.

[DanClemmensen]

A low-pressure gas-on-gas preburner to pump liquids into the thruster's main combustion chamber?

A low-pressure gas-on-gas burner to gassify LOX and methane?

Let me expand on this. If I need a bunch of high-pressure gas for my pressure-fed hot-gas thrusters, I should be able to produce it by heating LOX and liquid methane. I can create the heat I need for this by burning low-pressure ullage gas in a specialized burner (effectively a low-pressure hot-gas thruster), using the hot exhaust from this burner in heat exchangers to gassify more LOX and liquid methane. I fire up this system a few seconds before I need the high-pressure gas to pressurize the system and build a reserve, and I then run it to keep up with the demand from the hot-gas thrusters. Bleed a small percentage of this hot gas back into the main tanks to maintain the ullage pressure. Cooled burner exhaust is vented to the outside.

[Vultur]

Isn't one Raptor far too powerful to land on lunar surface. More than 200 tonnes of thrust verses 1/6th weight of Starship?

It's too powerful to hover on the lunar surface. But you don't need to hover to land. F9 doesn't.

[Robotbeat]

Could also just throttle low enough. Since you’re in vacuum, you don’t have to worry about nozzle problems from overexpansion at low throttle like you do on Earth.

[rsdavis9]

Could also just throttle low enough. Since you’re in vacuum, you don’t have to worry about nozzle problems from overexpansion at low throttle like you do on Earth.

musk mentioned that the limit was flameout in the preburners.

[OTV Booster]

The Shuttle did 104% IIRC.

I don't think that was overthrottling, it was in reference to efficiencies over the original engine design. Kind of like if Raptor V3's were rated relative to 100% of Raptor V1's thrust, doesn't mean the Raptor 3's are being run at "150% throttle".

Hmmm. I was about to ask if maybe our SD thrust numbers are current but they probably are. They're not used for Dragon landing as originally intended, and haven't run an escape sequence since their last test. Anything but unchanged would be a surprise.

[OTV Booster]

I see one way to find out for sure.

Land first with thrusters, drop a rover or two with cameras to watch the ship take off and then land again on the raptors.

I LIKE this. F9 has hoverslam down pat - on a prepared surface, in a 1g field where it learned the technique hoverslamming the ocean. If Arty III does hoverslam only and it doesn't work out, Arty IV becomes Arty IIIb. Not a way to move the program forward.

[OTV Booster]

I don't follow McGregor very closely but have a maybe unfounded impression that BC has much greater visual coverage than McGregor. Could small engine testing there slip through the cracks?

We saw SD testing there, didn't we?  This is of the same scale.

And I'm not sure about SD testing there. The engine is hypergolic and the exhaust is nasty.

I don't know how this gets handled for the D2.  I'm pretty sure there's an SD testing cell at McGregor.  I assume that SD's need to be qualified before they're installed on a new D2.  Does anybody know if they need periodic re-qualification?

Note that the backflow bug is a propellant supply system problem, not an engine problem per se.  I would assume that the test cell has a custom propellant supply system, and the MMH/NTO tanks are hardened against a combustion chamber or nozzle explosion.

On another point, settling thrusters may be in the range of 1kN but horsing around a bucking sloshing booster may take more oomph. Obviously cold gas works well with the booster flip and it may be fine to settle down a sloshing ship with a desire to get intimate with another ship. If everything gets oriented correctly and all that is needed is final translation, a set of 1kN thrusters and cold thrusters may be all they need.

Here's a maximalist view of the requirements, with a fairly maximalist mapping of those requirements onto four different kinds of thrusters, in addition to the Raptors, which obviously handle large delta-v maneuvers:

1) Small delta-v (5-50m/s) orbital maneuvers:  big (~70kN) thrusters, up to 100s burn time.  (I think Raptors are too big to get decent residual precision.  They'd do in a pinch, but if you have the 70kN thrusters, they're a better fit.)

2) Lunar and Mars landing:  big thrusters, up to 20s burn time. (More thrusters for Mars, but the lunar Starship weighs a lot more, due to the ascent prop mass).  Raptors seem unlikely for both, for the same reasons:  bad visibility, too much FOD, and landing site excavation.

3) Coarse attitude control:  When I worked this out for Starship 2, I got a combined pitch or yaw thrust of about 30kN from the nose and tail, if you want an angular acceleration of 0.5º/s². Maybe 5s burn time?  That seems like it's also a job for the big thrusters.

4) Fine attitude control:  <<1kN cold gas thrusters, up to 5s "burn" time.

5) Ullage acceleration:  <<1kN combusting gas thrusters.  Cumulative burn time might be more than 5000s, but it's possible they can be staggered to avoid overheating, or maybe even the whole system can be pulsed, with enough thrust decay that the prop doesn't rebound when they get shut off.

The ullage thrusters are a PITA.  If they could be combined with the big (70kN) thrusters, that'd be great.  But they have a substantially different operating mode from the big thrusters.

The other major question is whether the big thrusters and ullage thrusters need to be pump-fed or not.

IIRC, the D2 prop supply problem is a result of one set of tanks feeding both Dracos and SuperDracos. If hypergolics feed only the landing engines, I think the problem goes away.


Edit to add: the D2 kablooie was at the Cape during thruster testing, so at least the system level testing is done there. This says nothing about individual engine testing.

[OTV Booster]

You've been touting heated COPVs for thruster gas for quite a while. It's a good idea with only one possible change that suggests itself. Charge them up with autogen tap off and apply heat as necessary to keep the pressure up.

Even if you can cram enough gas into a COPV at the right temperature and pressure via the tap-offs, it'll still require lighting a Raptor to do it.  If you're in deep space and just want to do a TCM burn or an attitude change, you probably prefer not to do that for what's essentially a housekeeping function.

You don't need pressurant filled quickly; you just need it in the right state before you have to do something.

I think gas fed thrusters have the virtue of faster startup than liquid thrusters. That's one box ticked off on the wish list. I also suspect they can do deeper throttling. Probably much lower than 40%, but that's conjecture based on acetylene torches. Maybe another check box.

The problem with gas-fed thrusters is you have to store the gas somewhere, and then pV = nRT is not your friend.  You need the density of liquids to reduce the storage dry mass to something manageable.

If you can gasify the CH4 by running through the regen system, that's fine.  But liquid has a lot of advantages.

And of course I'm skeptical that pressure-fed will work.  If you have to go to pumps, then it has to be liquid at the inlet.

The visibility of early SD tests might be attributable to the nasty brown/orange smoke plume of hypergolics. With McGregor growing and getting busier than the old days maybe gas-gas has slipped through the cracks. They're stone simple and I doubt testing would be as involved as liquid engines.

The combustion products of MMH+NTO are about a third water vapor.  It may or may not be an orange cloud, but it's gonna be a cloud no matter what.

IMO, if they do use SDs for landing engines it would have to be a 'git er done' thing. It's two more fluids and that is high on the Musk list of sins. If gas-gas can be made to work it takes settling props out of the maneuvering equation. If gas-gas works for maneuvering it's a strong incentive to use it for landing engines too, but ya gotta do what ya gotta do. If/when lunar landings become routine there'll be landing pads and using raptors for landing starts to make sense.

I agree it's only temporary, as I've said several times.  But the lack of visible progress makes me think the temporary solution is all that's available.

3) Let's for a moment think that SpaceX/Elon hasn't given up on landing on Raptors alone.  It's simply impossible that their confidence in that approach is more than, say, 75%. 

Technical confidence, or confidence that NASA will allow it? There was a reference in that "Moon and beyond" post to Raptor/surface interaction tests; I'm still not ruling out the possibility that they're pretty much certain it will work.

Both.  (At the end of the day, there's not a huge difference.)

They absolutely need Raptor/surface interaction tests, because they need to know where they have to shut off the Raptors, which controls how much prop the thrusters need.  But that could be more than 100m above the surface.

I guess I can't rule the possibility out, but I'd put the probability of being able to land all the way on Raptors at <5%.  Remember:  5.2GW of jet power for one Raptor, and a hoverslam on a rough surface with bad PNT.  "No" seems like a no-brainer to me.

The point was to charge up or top up the COPVs during normal engine ops. Not a special burn. The MCC chamber pressure is 350bar and the injectors higher yet. Follow the path back and the highest pressure is at the main pumps. Probably in the 600-900bar range.

If the COPV's are good to 5000psi that's 340 bar. Uncontaminated O2 can come off the pump and store at 340bar. Methane too but without contamination issues. I've no idea if this is gas, liquid or supercritical gas. The question of letting it cool or keeping it heated depends on how soon it'll be needed. Many use cases here.

Something like this can feed hot and cold maneuvering engines. Maybe landing engines too but I think this bumps the storage volume enough to warrant a separate judgment. Long term storage, like on a mars mission, is yet another set of questions. We've been noodling specialized builds but shying away from a special build for deep space. Worth a rethink? Maybe.


Edit to add: My bad. What I was calling Arty III and IIIb would be Demo Landing I and II. What I called Arty IV would be III.