Hacker News Reader

Comment by carabiner | Hacker News Reader

So the struts (plus supporting structure) are lighter than the legs? Why is that?

> So the struts (plus supporting structure) are lighter than the legs? Why is that?

Besides the other answers you've received, the lugs hold the booster from (near) the top. This means that the body of the booster is in tension during and after landing. Legs, on the other hand, support the landing load and weight after loading in compression. The booster is basically a thin-shelled tube, which is limited in compression strength (for a given wall thickness) by buckling; in tension, the strength approaches the strength of the material, so less additional reinforcement is needed in the structure to support landing loads.

The tanks can surely be pressurized at landing, which greatly helps to avoid buckling.

Pressurized with what? They've already used their fuel for the landing. They can't put anything else in the tanks without worrying about contamination for the next flight.

Pressurization gases? The fuel goes from tanks to engines (engine pumps) because tanks are under pressure, right? Even if the liquids are spent - they are rarely spent in full - the gases remain.

Helium is a common pressure/purge gas in the fuel/oxidizer tanks.

They autogenously pressurize the tanks - they heat up the cryogenic propellants with the engines and use some of the gas to pressurize the tanks. In Starship’s case it’s methane and oxygen.

But why, if you don't need to

Aghm, sorry, I meant, the tanks actually are pressurized at landing and not at risk of buckling. Why depressurize?

The pressure is enough to help push out liquid fuel but I don't think that means the pressure can be fully relied upon to provide structural support.

That's literally how original Atlas rockets and modern Centaur stages work.

Starship is literally not an Atlas rocket or Centaur stage.

Starship uses autogenous pressurization, which is not what Atlas/Centaur used.

Sure, but I am quite sure pressure is expected to be in place & provides the necessary strength for all the maneuvers.

You know, that's completely unimportant. The important parts are that 1) Starship stage is under pressure when landing and 2) pressurization makes a thin-walled metal cylinder much stronger resisting buckling. Details of how Starship works and how pressurization is historically used to increase strength are just to support these two points. But if you already have these two points, you should admit that the argument "Starship can't land on legs because there's too big of a risk of buckling" has some counterarguments. And the overall decision isn't as clear as we'd like to have it.

Actually, it is very important. Autogenous pressurization has a much higher risk of pressurization loss than a system which uses inert gases, due to the potential for the ullage gases to mix with the liquid fuel and condense.

This is relevant when designing the landing system.

> surely

Rewrite:

"Why doesn't [huge successful project] do [simple thing]?"

At least link to some details of the design? Here's the best diagram of the tank design I could find:

https://www.elonx.net/wp-content/uploads/SpaceX-BFR-spaceshi...

Which doesn't show the design constraints but who wants those - edit and it's not an image of the booster? Elon mentions a design feature missing from the diagram: https://x.com/elonmusk/status/1093643894917492736 I would personally guess you'd need to be very careful with your implied load bearing connections between the tanks at x Kelvin and the skin at redhot reentry temperatures...

Good luck on buying spaceY.com and competing against those engineering fools at SpaceX ;)

I am mocking unreasonably, and I know I would find similar comments in my own internet history. I am hoping you will learn to be a little less thoughtless in your armchair. We all assume other rocket-science engineers must not know what they are doing but usually that just shows our own ignorance.

Nah, misses. We discuss technical possibilities, not flame on forums. The previous post was an answer itself.

Or you're implying that tank pressurization isn't a standard practice and not a simple thing?..

Let me explain once more :) . The original post which I was replying to was

----- > So the struts (plus supporting structure) are lighter than the legs? Why is that?

Note how the author says that a thin-shelled tube is limited in compression strength by buckling. Technically it's correct, but practically if you put some extra pressure in that tube - which, after all, has also airtight caps on both ends - then the tube becomes much stronger, and is able to withstand reasonable forces during landing.

That's what I noted, and I can repeat that. I am quite sure SpaceX engineers considered that possibility, and I think they rejected that because they felt they see an even better result. I'm trying to see that here.

I also suspect that you don't know my qualifications in the area, and referring to armchair ones just so. It's interesting how many different and widely qualified people participate in HN discussions.

What is the booster’s body in during launch/flight?

Tension from internal pressure -- these aren't really balloon tanks, but they absolutely benefit from internal pressure.

What about during the slow-down burn?

The same, its just a much higher proportion of gas rather then liquid. Basically on the pad its mostly full with liquid, as it launches, it pumps back part of the gas created in the engine back into the tank. That called 'Autogenous pressurization'. So they don't need an extra gas like helium, as for example Falcon 9 needs.

The big issue during landing is that you need to make sure that the engine doesn't suck in gas. That causes bubbles and can destroy the engine. This was actually the failure that caused some of the earlier SN flights to explode or not produce enough power from the engine.

You need to either have header tanks, like the booster. Or some kind of method to push the liquids into the right place.

If you want to deep dive into the whole problem, 'CSI Starbase' on youtube has a brilliant series on all the engineering problems with all of this. Its a very complex problem.

Ok, so I was right: Does this end up being more efficient because more of the loading is in tension instead of compression?

This bugged me because everyone was saying the deletion of legs was key, but to me the struts are basically legs mounted up high. It's taking advantage of tensile loading that promotes the weight reduction.

the other big difference is that legs need to extend below the engine which means they need up move, which makes them much bigger and more complicated than the catch pins

Also, the struts are much smaller than the legs, and there's no need for moving parts or hydraulic mechanisms.

The booster is already strong enough to support itself in compression, because that's what it does during ascent and the landing burn. The entire bottom structure of a rocket (the "octaweb" for F9) is basically made to transfer the thrust compression loads of the engines into the tanks.

The legs would need to be much longer (because you can’t push the engine nozzles all the way into the ground and still hope for good things.) Longer structure means more mass, but also larger torque which need to be handled with the support of the structure.

Legs need to move to deploy. The struts are just there, static things are much simpler. Simpler things weigh less.

Legs need to contain shock absorbers. With the struct solution the shock absorber is in the chopsticks. It doesn’t matter how much the shock absorber weighs when you don’t need to carry it up with you.

It was pretty neat how Bezos explained to Everyday Astronaut how they were using 6 legs so the legs weren’t as long.

On the whole, pad catch is the way to go for non-expeditionary vehicles. For orbital uses each booster basically becomes a pyrotechnic elevator.

It would be good to see the numeric analysis of variants here. Legged landings are surely possible - say, with longer legs (twice as long as struts?), possibly static (legs are always deployed, even when launching), with shock absorbers which aren't that heavy... Would be good to see good and bad qualities next to each other.

I think it all just comes down to weight. If you can trim mass on the vehicle then you should do it.

The pins/struts are a 2 point system that double as the booster lift points in general operations. The booster mostly hangs in tension which the existing tank structure can support. I would guess they share some of the structure beefiness with the grid fins.

Legs require at least 4 points, probably more. Shock absorption hardware, ability to unfurl to an acceptable width. Require reinforcement (cross bracing) near the base of the tanks to handle the loads pushing inwards toward the center of the tanks.

Legs require at least 3, not 4 points.

You can technically imagine two legs with really wide feet, allowing some perpendicular stability. I wonder if one-leg lander could be imagined. 3-legged landing scheme was used in Surveyors, first American automatic Moon landers, and was surely considered for Appolo LEMs, but rejected. So there could be additional, secondary reasons when choosing the number of legs.

The cost of one additional leg is pretty inexpensive for the redundancy it provides for the other three.

4 legs have no additional redundancy over 3. One leg failing will still result in the booster tipping over. They do push the maximum angle of tipping before your CG is no longer supported out farther though.

Useful for when you land on the rim of a crater like Apollo 12.

I think you are both using different meanings for the word 'least'.

Nitpick, I think the issue is different meanings for the word ‘require’. Practical/realistic minimum vs absolute/literal minimum.

The "struts" are needed anyway to lift/move the rocket using cranes, they aren't optional.

They’re much smaller and don’t have to move/deploy.

Reply on news.ycombinator.com