Wiki Super Heavy/Starship - General Development Discussion

[JB47394]

Hmm... aren't we talking small RCS thrusters here? How would the Raptor thrust imping on the other ship?

In general, if you were to design an approach for docking between two vehicles, how would you do it? I'd do it tangentially because the head-on approach has no redeeming characteristics other than being able to squeeze small vehicles into the complex structure of a vehicle like the ISS.

I wasn't trying to suggest that Raptor thrust was going to impinge the other ship on a Starship refueling. That's why I mentioned battleships and space stations - things that are inordinately beefy and would require huge thrusters. But do you want even small thrusters hitting your vehicle? Perhaps pushing exhaust gases into nooks and crannies, or across your optics or electronics or sensors, or knocking loose a thermal blanket? Think about the waste products in rocket exhaust.

That eliminates the need to use thrust to move towards each other... gotcha.

Using thrust to move towards the target vehicle isn't a terrible thing. The exhaust from your engines isn't striking the target vehicle - so long as you only accelerate towards the target vehicle.

Dragon approaches the ISS very slowly and from a long distance because of the head-on approach. They're on a collision course with the ISS from the start, so they have to come in slowly. Compare that with a Dragon that comes in tangentially. They could come in at 20 m/s, and fire heavy thrusters to brake a few meters from the docking port. Then they could tiptoe up to soft capture. Just make sure that they only accelerate towards the station.

Shoot, maybe we should just use ropes like we used to throw between sailing ships to draw them together.

 

[TunaBug]

Okay but that won’t lead to the ship having a full tank. The tanker ship will never arrive in orbit with a full tank. So how does the ship needing prop ever get full if no pumps are used?

The fuel to be transferred (CH4 & LOX) are cargo, and I have always assumed that the tanker ship would be a version of Starship where the cargo area was occupied by tanks.

That also solves the problem of simply letting "gravity" (acceleration from thrust in this case), cause the levels to equalize: the cargo tanks supplying the fuel would be "higher than" the tanks receiving the fuel, so the tendency for the levels to equalize would leave the cargo tanks empty.

To be honest, I hadn't even thought of what I just wrote when I made the previous post, which was definitely a case of talking before thinking. So I took your question as a dare and came up with something, still expecting somebody is going to say "this is dumb because...".

 

[TunaBug]

I didn't get that impression for the afore-mentioned Ars article:



They seem to make the distinction that the thrusters are for settling the prop, and then use "fine tuned tank pressures" for the prop flow...

What you wrote is your interpretation of what you quoted, and it fits with what you quoted. But they didn't supply that detail, you assumed it. If you simply don't turn off the engines (and they never said the engines would shut off), then the pressure differential can come from thrust. Just a different interpretation from a different imagination filling in the blanks.

And as I wrote above, I was definitely posting before thinking. But I don't see that my imagination contradicts the part you quoted any more than your imagination.

 

[scaesare]

Yeah we had some subsequent discussion and it's hard to make out the actual head design.. if they are rounded all the way or there's and edge/barb on the bottom. Even if it is the latter, I wonder if having the tile slot edges bent inward a bit to present more of an edge to capture the pin barbs would help...

(responding to my own post)

Some interesting discussion/speculation about the tile fastening/pin problems:

 

[scaesare]

What you wrote is your interpretation of what you quoted, and it fits with what you quoted. But they didn't supply that detail, you assumed it. If you simply don't turn off the engines (and they never said the engines would shut off), then the pressure differential can come from thrust. Just a different interpretation from a different imagination filling in the blanks.

And as I wrote above, I was definitely posting before thinking. But I don't see that my imagination contradicts the part you quoted any more than your imagination.

But thrusters don't change the tank pressure, right? The settle the liquid to one side to cover the intake ports, but they don't change the pressure. Yet that's what they say will cause the liquid to flow.

 

[Ben W]

But thrusters don't change the tank pressure, right? The settle the liquid to one side to cover the intake ports, but they don't change the pressure. Yet that's what they say will cause the liquid to flow.

Hot gas / ullage thrusters do, because they use the boiloff gas in the receiving tank (reducing the tank pressure in the process) instead of the liquid fuel.

I still think that eventually the transfer would best be done using centripetal force to settle the propellants, which would eliminate the need for continuous thrust during the transfer. After docking, the tanker can use its boiloff gas to initiate the spin, which will also initiate the pressure differential. If more differential is needed, it can be created by actively pumping boiloff gas (not liquid fuel) from depot to tanker, causing the liquid propellant to passively flow from tanker to depot. This would further reduce unnecessary venting / wasting of propellant. When the transfer is complete, the tanker can use its own higher-pressure boiloff gas to spin down the paired ships before disconnecting. If the ships have effective sunshades (particularly the depot), it's possible that some of the boiloff could then be re-condensed, rather than vented/wasted. In space, every kilogram counts!

 

[TunaBug]

But thrusters don't change the tank pressure, right? The settle the liquid to one side to cover the intake ports, but they don't change the pressure. Yet that's what they say will cause the liquid to flow.

If you keep thrusting there is downward pressure, the exact same pressure that you're talking about using to settle the tank. Perhaps I'm abusing the term "pressure": so be it, but I cannot imagine that anybody writing these articles is that precise and correct in their use of terminology, and that's even assuming they even know precisely how things will work.

Here's a question that I do not know the answer to: If they thrust to settle the liquid, then remove the thrust while the liquid starts to flow (and using a pressurized tank as you describe), is surface tension sufficient to keep the liquid at the exit while in microgravity? Or will they need thrust continuously anyhow?

 

[JB47394]

still expecting somebody is going to say "this is dumb because...".

I wonder which way they'll go with a tanker. Will they treat the propellant like cargo or will they stretch the tanks proportionately into the cargo area? Treating it like cargo means that everything remains the same, except that there needs to be plumbing and valves to direct propellant to and from the cargo area via the quick disconnect ports. They might be able to leverage the plumbing that goes to the header tanks. It would also mean another bulkhead to separate the methane and oxygen in the cargo area. Stretching the tanks only means a longer downcomer. (Which apparently is not the word that SpaceX uses)

If you simply don't turn off the engines (and they never said the engines would shut off), then the pressure differential can come from thrust.

Running one Raptor sea level at 40% thrust (the lowest thrust level they can produce) generates 92 tons of thrust. Assuming a 1500 ton combined mass, that's 0.6 m/s^2 acceleration. If the transfer takes an hour, that's 2,160 m/s of delta-V. Not to mention a whole bunch of propellant gone. Cut it in half and it's still a boatload of delta-V and propellant mass. Note that loading Starship on the pad takes something like half an hour, and that's with pumps.

In the NSF video, Jack states that the vehicles will use "settling thrusters", which is not the way anyone would refer to Raptor engines. On the slide they show, there is mention of the need for "hot gas thrusters", which will probably be the completion of the methalox thrusters that they were working on a while back.

is surface tension sufficient to keep the liquid at the exit while in microgravity?

Liquid oxygen has really low surface tension. It's less than 20% that of water.

After docking, the tanker can use its boiloff gas to initiate the spin, which will also initiate the pressure differential.

Go nose to nose and mechanically spin up in opposite directions. It's probably dynamically unstable, but it eliminates the need to vent anything.

 

[Ben W]

Go nose to nose and mechanically spin up in opposite directions. It's probably dynamically unstable, but it eliminates the need to vent anything.

I pictured the usual side-by-side mating configuration, with the spin axis where the vehicles are touching, mutually spinning around each other with the noses staying pointed a fixed direction. This should be a stable spin axis, and if the noses are pointed into the sun, a small sunshade could protect both vehicles. The gas transfer line could be very short, with a longer liquid transfer line connecting the far sides of each tank relative to the spin axis. Since constant thrust is not needed in this scenario, the propellant transfer could be done arbitrarily slowly (with smaller pipes and pumps) without wasting fuel.

 

[JB47394]

I pictured the usual side-by-side mating configuration, with the spin axis where the vehicles are touching, mutually spinning around each other with the noses staying pointed a fixed direction. This should be a stable spin axis, and if the noses are pointed into the sun, a small sunshade could protect both vehicles. The gas transfer line could be very short, with a longer liquid transfer line connecting the far sides of each tank relative to the spin axis. Since constant thrust is not needed in this scenario, the propellant transfer could be done arbitrarily slowly (with smaller pipes and pumps) without wasting fuel.

Yeah, I understood. In thinking about it more, I wonder how stable that would be. The propellant would slosh, the center of mass would move around during transfer, and even the location of the propellant would move. They'd probably want radial baffles in the tank to direct the propellant.

Hmmm. If the baffles are corkscrew shaped, then the nose-to-nose configuration could drive the propellant towards the nose on the tanker and away from the nose on the receiving ship. The ships themselves would operate like giant pumps.

Hmmm. Connect the ships, and then connect their internal impellers so they spin opposite to each other. The ships themselves stay stationary.

Note to self: Call Elon.

 

[Ben W]

Yeah, I understood. In thinking about it more, I wonder how stable that would be. The propellant would slosh, the center of mass would move around during transfer, and even the location of the propellant would move. They'd probably want radial baffles in the tank to direct the propellant.

Hmmm. If the baffles are corkscrew shaped, then the nose-to-nose configuration could drive the propellant towards the nose on the tanker and away from the nose on the receiving ship. The ships themselves would operate like giant pumps.

Hmmm. Connect the ships, and then connect their internal impellers so they spin opposite to each other. The ships themselves stay stationary.

Note to self: Call Elon.

The spin-up would be done gradually, and the propellant would quickly reach an equilibrium position along the outer tank walls. The existing baffles (or just friction with the tank walls) would quickly absorb the energy of any slosh, and without any active energy inputs (after the spin-up is complete) it would settle into a slosh-free static equilibrium after just a few minutes. This is completely different from the highly dynamic environment of hot-staging, or Starship’s landing flip maneuver.

 

[JB47394]

Road closure tomorrow from 7AM to 7PM CDT, possibly to move S30 to the suborbital pad for a static fire.

It turns out they were doing some testing of ground support equipment. Certainly they've completed a huge amount of work on the launch site so far. Busy as beavers.

 

[Grendal]

SpaceX making progress on Starship in-space refueling technologies

NASA says SpaceX is on track to demonstrate in-space refueling of Starship next year, a critical technology for returning humans to the lunar surface.

spacenews.com

 

[scaesare]

Hot gas / ullage thrusters do, because they use the boiloff gas in the receiving tank (reducing the tank pressure in the process) instead of the liquid fuel.

I still think that eventually the transfer would best be done using centripetal force to settle the propellants, which would eliminate the need for continuous thrust during the transfer. After docking, the tanker can use its boiloff gas to initiate the spin, which will also initiate the pressure differential. If more differential is needed, it can be created by actively pumping boiloff gas (not liquid fuel) from depot to tanker, causing the liquid propellant to passively flow from tanker to depot. This would further reduce unnecessary venting / wasting of propellant. When the transfer is complete, the tanker can use its own higher-pressure boiloff gas to spin down the paired ships before disconnecting. If the ships have effective sunshades (particularly the depot), it's possible that some of the boiloff could then be re-condensed, rather than vented/wasted. In space, every kilogram counts!

Hmm.... good point about hot gas thrusters effectively lessening the tank pressure as a byproduct of their usage... I wonder if that's the mechanism they are using to manage pressure in addition to settling the tanks... I wouldn't have assumed so, but I suppose you could fire opposite thrusters simultaneously if you wanted to drop pressure yet not impart net thrust to the ship...

 

[scaesare]

If you keep thrusting there is downward pressure, the exact same pressure that you're talking about using to settle the tank. Perhaps I'm abusing the term "pressure": so be it, but I cannot imagine that anybody writing these articles is that precise and correct in their use of terminology, and that's even assuming they even know precisely how things will work.

I assume when they are saying "SpaceX will fine-tune tank pressures", that isn't what they are talking about, because the inertial pressure the liquid exerts on one surface as it's being settled doesn't change the overall tank pressure.

But that does indeed imply the terms are being used accurately....

Here's a question that I do not know the answer to: If they thrust to settle the liquid, then remove the thrust while the liquid starts to flow (and using a pressurized tank as you describe), is surface tension sufficient to keep the liquid at the exit while in microgravity? Or will they need thrust continuously anyhow?

(on edit)

Good question. Once the valve opens, causing a low pressure region for the liquid to flow, does the pressure differential on one side of the liquid blob vs the other keep it together, assuming it's not agitated significantly? I'd guess surface tension does indeed play some part there...

 

[JB47394]

I wouldn't have assumed so, but I suppose you could fire opposite thruster simultaneously if you wanted to drop pressure yet not impart net thrust to the ship...

Fire the settling thrusters on the receiving ship through the center of mass of the combined vehicles. The receiving ship's tanks drop in pressure, pulling propellant from the tanker. Then it's a matter of maintaining pressure on the tanker side. They might be able to do that by leaving the tanker in the sun. It could even produce so much boiloff that they'd actually vent to keep pressures in the right range. Given that they have heat tiles on one side and steel on the other, they should be able to tune the tanker's thermal soak, keeping venting down.

does the pressure differential on one side of the liquid blob vs the other keep it together

Imagine that you're in space and there's a blob of water floating in front of you that you're going to drink from using a straw. As you draw water down the straw, it's going to cause the blob to collapse onto the end of the straw. The gas pressure around the blob does that. But it's the surface tension of the water that keeps the blob together. Without that strong surface tension, the blob would deform, and the gas around the blob would find a path to the straw - because it's the gas pressure that's pushing the blob towards the straw.

Liquid oxygen has very low surface tension, so once the liquid starts going through the straw, the surface of the blob is going to more easily deform. In other words, it'll be easier for the gas to intrude into the liquid and reach the straw. I'd need a simulation to see how serious that effect is, but you can see the interior of liquid oxygen tanks when they hit zero g and there are little globules all over the place. The stuff doesn't really care to blob up that much.

 

[JB47394]

Road closure tomorrow from 7AM to 7PM CDT, possibly to move S30 to the suborbital pad for a static fire.

You were only a skosh off. Ship 30 was moved to the suborbital pad early this morning. This may be the last ship to static fire at the launch site.

 

[TunaBug]

I wonder which way they'll go with a tanker. Will they treat the propellant like cargo or will they stretch the tanks proportionately into the cargo area? Treating it like cargo means that everything remains the same, except that there needs to be plumbing and valves to direct propellant to and from the cargo area via the quick disconnect ports. They might be able to leverage the plumbing that goes to the header tanks. It would also mean another bulkhead to separate the methane and oxygen in the cargo area. Stretching the tanks only means a longer downcomer. (Which apparently is not the word that SpaceX uses)


Running one Raptor sea level at 40% thrust (the lowest thrust level they can produce) generates 92 tons of thrust. Assuming a 1500 ton combined mass, that's 0.6 m/s^2 acceleration. If the transfer takes an hour, that's 2,160 m/s of delta-V. Not to mention a whole bunch of propellant gone. Cut it in half and it's still a boatload of delta-V and propellant mass. Note that loading Starship on the pad takes something like half an hour, and that's with pumps.

In the NSF video, Jack states that the vehicles will use "settling thrusters", which is not the way anyone would refer to Raptor engines. On the slide they show, there is mention of the need for "hot gas thrusters", which will probably be the completion of the methalox thrusters that they were working on a while back.


Liquid oxygen has really low surface tension. It's less than 20% that of water.


Go nose to nose and mechanically spin up in opposite directions. It's probably dynamically unstable, but it eliminates the need to vent anything.

(on edit)

Good question. Once the valve opens, causing a low pressure region for the liquid to flow, does the pressure differential on one side of the liquid blob vs the other keep it together, assuming it's not agitated significantly? I'd guess surface tension does indeed play some part there...

I don't recall that SpaceX ever provided details, or even an overview, on the propellant transfer demo that they did in IFT-3 as far a "how it works". I think I just a assumed a pump between two tanks and thought it was a stupid demo. Now I'm realizing that it was a silly interpretation on my part.

It's sounding to me, at this point, that they may need continuous force (thrust, centrifugal, whatever) to keep the liquid at the ports, as well as pressurizing with a gas (from wherever) in order for things not to take forever. Sounds like a lot of neat problems. And probably something that they don't want to disclose.