Hacker News Reader

Comment by modeless | Hacker News Reader

The booster is destroyed when it tips over after "landing" vertically on the water. It's like a 20 story building falling over.

If you're asking why they don't land it on a floating barge like Falcon 9, there are two reasons. One is that landing back on the launchpad lets them refuel and relaunch immediately. The other is that landing legs are big and heavy and significantly reduce the payload capacity. If you're already landing on the launchpad you might as well add arms to catch it. The mass of the arms is free because they're on the tower instead of the rocket, and the rocket only needs tiny nubs to catch on the arms instead of giant legs. Also the arms double as a crane to lift and stack the rocket on the launchpad.

View on HN

Like a falling chimney (google it), it breaks because the top can't fall fast enough.

Also, a rocket that hangs doesn’t need near the structural beef as one that is designed to withstand landing on its feet.

Not convinced about this argument; wouldn't the forces during the slow-down burn be applied much the same as the from the feet?

Actually no - the forces travel inside the "tank" of the rocket, and push against the top of it.

The force doesn't push against the nozzle of the engine.

Same pressure in the tank as in the nozzle? That doesn’t sound right. It would burst.

Same force, not same pressure. The nozzles are relatively small.

I wouldn't want to try to intuit what the force distribution would be and how much is carried through each component of the structure, though — that's what simulations are for.

> The force doesn't push against the nozzle of the engine.

I’m not sure what you are saying here. The force pushes against the nozzle, (and of course the walls of the combustion chamber.) That is the purpose of the rocket engine, to push the rocket forward. They are not just there to provide mood-lights.

I said "nozzle", as in the side walls of the engine.

The force doesn't transfer via those walls, rather it goes through the full tank of the rocket and pushes against the top.

> I said "nozzle", as in the side walls of the engine.

Ok? I seem to have said the same.

> The force doesn't transfer via those walls

Where do you think the force starts on a rocket? What pushes the rocket forward?

The chopsticks on the tower can absorb a lot of the impact, I'd guess.

Also, a ship big enough to carry Starship booster would be huge. And then what port infrastructure do you use to bring it to land and transport it to your launch pad? You would basically need your own port right next to the launch site.

Are they going to need legs when they land the Starship on the moon or Mars?

the booster does not need legs, only starship does

Yes, also the lunar variant won't need a heat shield so that will compensate for the extra leg mass. Not sure what their plans are for a heat shield for Mars.

They're planning on catching starship as well however, as the in-orbit refueling will require a lot of starship launches that aren't going anywhere except orbit and back.

Yes, Starship needs legs to land on the Moon or Mars. They will be lighter than legs for Earth landings, as Moon/Mars have lower gravity.

Not sure about the amount of fuel needed to land, as there is much less atmosphere on Mars, and none on the Moon. I would still guess that it needs less fuel there as well.

Don't they also need fuel to lift off again?

I think Elon has said that it's going to be a one way trip to Mars.

That will have to happen through in-situ resource utilization. aka making the fuel on Mars.

An atmosphere reduces the amount of fuel needed to land, because you can use aerobraking to slow down rather than carry fuel to do it. See the Apollo return capsule, which landed without any rockets, only parachutes and a heat shield.

Yes. Starship HLS will have them.

But the other Starships don't and adding them would make a big difference.

Since you need a huge fuel tank in orbit that needs to be refilled by multiple starships to refuel Starship HLS to actually land on the moon adding legs on those refueling Starships would decrease payload capacity, and thus you'd need even more Starships to refuel the tank for a single lunar landing.

> One is that landing back on the launchpad lets them refuel and relaunch immediately.

How do they structurally, electrically, etc. checkout the rocket after landing?

An ops problem for later. At least now its possible!

It seems like a pretty expensive assumption that landing the rocket is enough. Relaunching the rocket, which requires inspection and validation procedures and technology, is just as important as landing it.

They haven't done it yet and nobody outside of SpaceX knows their specific plans. Of course we can speculate that they will design sensors and cameras to replace any previously required manual inspections.

>immediately

So far this hasn't been shown even on much simpler Falcons. The barge gives quite some energy advantage for not having to boost back.

Ideally they'd build a capesize kinda barge with chopsticks to catch it in the ocean, then perhaps service what they have to while it's steaming back.

Hasn't been shown because they're not meant to be immediately reusable.

Spaceship and/or super heavy are (I don't remember the details).

I'll believe it when I see it.

It would be good because there are many people who, roughly, refuse to believe their own eyes and keep moving goalposts, misrepresenting what they said or meant earlier, inventing additional conditions, changing their mind etc. If you would honestly believe that Starship can fly frequently when you see it flying frequently, you're already ahead of some.

I remember when people doubted that the full-flow staged combustion methalox engines could work... until SpaceX showed over a hundred of them working now.

Then the ULA CEO Tory Bruno claiming that the SpaceX photos of the first Raptor 3 were “partially assembled”, to which Gwynne Shotwell replied with this: https://x.com/Gwynne_Shotwell/status/1821674726885924923?t=v...

I'd consider that cheating on the Gwynne side. It could be argued if chamber tilting mechanism is part of the engine - after all, it was added separately to NK-33 - but surely for Raptors it's considered an integral part, the fire test was then a test of chamber, not the engine :) . Though Gwynne had to answer positively, having little good choice.

The outer ring of engines on Superheavy do not tilt. They're rigidly mounted and exclude even the engine start hardware which is contained in the launch mount, further reducing weight. And they're only used during liftoff, not for boostback or landing.

Raptor is non-functional without tilting. That is, it's possible to build rocket where all Raptors are fixed, but that defeats some ideas of Raptors. And I haven't seen tilt mechanisms considered parts of the rocket.

So, while engines without tilting have been used - e.g. NK-15 - Raptors aren't from that category.

Raptor is used in a non-tilting configuration on both the booster and ship. You could certainly design a rocket to get to space with only non-tilting raptors, if you wanted. It would be silly to consider it incomplete without tilt, even if Starship does have some tilting ones too. Consider it two variants if you like, both complete on their own.

Differential thrust has been used to steer rockets for decades, and raptor has the sort of deep throttling capability required for it. Tilting is absolutely not required for anything but landing.

Yeah - you would still need roll-control, but given there are huge potato smashers already bolted on, it should not be an issue while in enough atmosphere. ;-)

Yup. In addition to the grid fins, there is the warm gas RCS which could perform roll control.

I'd like to suggest the idea that had they been able to operate it "well" without tilting, they wouldn't have added the mass for it.

There's clearly a huge advantage somewhere.

Yeah, landing. As was mentioned.

Not necessary for launch.

Also catching a booster, also landing boosters at all, also achieving cost reduction via landings, also cheap enough phased arrays to make Starlink viable, also also also... SpaceX has a long history of proving doubters wrong.

I remember talking about landing booster when I was in college in the 1980s. Super awesome I got to see it in my lifetime!

And yeah, Starlink was a great purchase by SpaceX.

What part of Starlink was purchased?

It’s fake news, SpaceX built Starlink internally. Even Wikipedia says so: https://en.m.wikipedia.org/wiki/Starlink

Not every piece of misinformation floating around is “fake news”

There's people on the Internet always arguing SpaceX was purchased. Starlink as an idea within SpaceX is older than it's first launch

Reread my comment.

I have but it's a spreading meme. Usually mixed with other falsehoods

Falcon 9 wasn't designed for immediate reuse. Immediate reuse is just a theory right now, but SpaceX has an excellent track record of turning their theories into reality.

SpaceX did purchase some oil rigs with the intention to turn them into launch platforms, but later abandoned the idea. It's probably something they will return to later once Starship is flying regularly. You're right that avoiding the boostback burn is a big advantage. But maybe they don't need to bring the booster back after it lands on a platform, it can just launch again from there. Maybe they could have a bucket brigade of launch platforms ringing the Earth!

The oil rig idea is brilliant, especially for cargo flights.

I.e. the bulk of near / intermediate term launches

Far fewer people care if you "oops" an unmanned rig in the ocean.

The issue is logistics, and in this case isnt an easy solve.

You have to get fuel, and the rocket, out to a pad in the ocean, and have to deal with a rocket lift on varying conditions.

If you dont want to do most of that, then the only option is putting your manufacturing on the rig too, which negates lifting a rocket, but instead makes the rig huge, and requires having a train of ships in and out 24/7 to keep it supplied.

There was even a company in the late 90s that tried oil rig launch platforms and ultimately abandoned it.

Most heavy lift is transported by barge over water anyway.

Since the 2 stages of Starship aren't intended to be road mobile, due to size, there's no transportation benefit to being land-accessible.

So really the main concern is piping propellant... but afaik some rigs off the shallow coast of Texas are directly piped to land?

The main benefit you get is terrifying the FAA et al. less, as the consequences of a missed catch are now out in the ocean.

Musk spoke of 24-hour Falcon turnaround as early as 2011 and as late as 2019.

Immediate reuse means significantly less than 24 hours. Falcon 9's current cadence is fast enough to meet their current launch demand and doesn't need to improve, especially with Starship on the horizon. On the other hand, Starship will need to launch repeatedly in a short amount of time for orbital refueling to work.

So it was declared as a target, was possible but wasn't done because no demand? Knowing SpaceX they'd do it just for bragging.

The fact that it's required for Artemis to work, and the amount (nobody knows exactly but lower bound is like 15) of Starships required to launch in quick succession just highlights how risky, to the point of unsoundness, the project is.

They don’t need rapid reuse for Artemis. It would certainly help but Starship can just hang out in orbit for a few weeks while they do what they need to do to launch all their rockets.

I wonder what the boil-off rate is on the cryogenic propellants.

It really depends - you can do direct injection to GEO now with cryogenic stages. And IIRC Soviets did some tests with kerolox stage (that should eventually launch on the ill fated N1) around the Moon, meaning multi-day flight times with liquid oxygen on board.

If you do the thermal design right, possibly use a sun shade (space is a large thermos bottle after all) or even use active cooling to remove the little heat that gets through, then it should work just fine. :)

The heat shield probably provides some degree of insulation from the solar radiation as well, and it's only needed on one side if you're far from Earth.

Yeah, I was thinking about the orientation during the coast phase & how thermal management could explain the Starship orientation at that point. :)

So is it cheaper to dump the booster in the ocean then land it back at the tower and then have to dispose of it?

The booster always follows a trajectory for an ocean landing until it passes a bunch of safety checks, and then it diverts to the launchpad. This way a failure early on can't cause it to crash on land.

What probably happened is it failed a safety check (e.g. a sensor read out of range) and so it didn't divert to the launchpad. It's cheaper to dump the booster in the ocean than to build a new launch tower if it's destroyed in a failed landing. They have an assembly line for boosters, but only one fully complete launch tower at the moment.

The launch tower takes multiple months to build. The orbital launch mount and launch tower and quick disconnect hardware is all custom and expensive and huge. They call it “Stage 0”. Losing that is an order of magnitude more of a setback than losing a Stage 1.

I guess they didn't optimize the towers to be fast-buildable :) because there are several obvious ideas.

The 2nd tower at Starbase is much more modular than the first and is being built at quite a rapid rate.

They’re building one at the Cape, too, but they’re still a multi-month process.

Construct additional pylons?

Why not to use the Mechazilla for adding the next section of tower on top of existing one? So we transport parts to the tower, assemble relatively thin sections of tower nearby, the lifter moving along the tower brings the sections on top of existing ones, and we have a few places for optimizations here. The proponents of The Boring Company may see some parallels here.

have they tried a giant net? lol

Yes: https://youtu.be/oTH3mq7SsK4?t=14

Reply on news.ycombinator.com