sevenperforce

I am pretty sure I remember Elon specifically saying that going with stainless steel means you need no leeward shielding or insulation at all.

The mounting points for the leg (for the F9 style, the telescoping cylinder) would have to pass through the interior heat shield. You would need a flush fit during re-entry or you would have re-entry plume recirculation between the carbon-carbon aeroshell and the inner tile pattern, in precisely the place where the machinery of the legs is stored. If you have a flush fit, you don't need the underlying tile pattern.

I mean I did say that the obvious fix was to use auxiliary gaseous methane tanks fed from the autogen press lines to supplement header tank pressure, used both for the header tank press and eventually for hot-gas thrusters. If only he had listened.......

I don't think they've made any actual engine choice yet so putting four engines down below is just notional. Obviously if they do put four engines down there, there's no way they can deep throttle enough to land on two of them, so they'd need a Rutherford (or maybe two or three) to perform the landing burn.

Looks like Elon had been playing stock KSP and was surprised by the switch to RO. Thinking about throttle settings. With two engines that’s definitely a suicide burn. But not all suicide burns are created equal. I am guessing that the flip will be timed such that the suicide burn will be achieved at or near minimum throttle for both engines, so if one has a thrust shortfall the other can throttle up to compensate. This is really the inherent problem with these landings...all this stuff is literally rocket science. Having three engines for redundancy is great because it allows these options but that doesn’t necessarily fix problems with timing. If something goes bad on a plane, you typically have plenty of time to come up with a solution, and there’s no mission-critical system that undergoes a bunch of transients right at touchdown.

It's a surprisingly compact solution. Like I said, I was able to do it in actual KSP without any clipping. But trying to move them inside the skirt is actually really tough. There are enough internal parts to make it a challenging fit. And you lose more of the wide stance than you would think.

So I actually went back to the drawing board and rebuilt this in KSP with the proper dimensions (using 3.5 meter parts instead of 5 meter parts) and.... ......I cannot believe how well it works. Like, it works REALLY well. Amazing shock absorption. I put a single Vector underneath and was able to do a 150 meter hop perfectly the first time with no problem whatsoever.

There might be enough space for six. One thing is that with this wide stance, though, you'd have to have a complete failure to not have SOME kind of support on that end.

Oh, it is clearly way, way too large. No question about that. I was just trying to design it in KSP which is inherently difficult since I don't have Tweakscale installed. Using active hydraulic control for auto-leveling allows shock absorbers to also be the auto-levelers. The best part is no part.

The current landing legs fold out onto the same points that are used to hold onto Starship while it is sitting on the pad. Presumably if you split the heat shield like this you would make the stowed-position feet themselves the hold-down points. I don't know if they could manage six or not. Four seems like plenty.

Okay, so this is definitely my favorite Starship leg concept yet. More:

I know that, and you know that, and @tater knows that, but apparently neither Elon nor Gwynne know that, so we have to assume they will make decisions based on that.

Yeah, if all these insane plans work and "propellant costs only" is really a thing, then you can totally have an abort-capable Starship separate from your BLEO Starship. But it won't work for P2P.

It was the kaboom that launched it in the air, not the 3 bar of ullage pressure. The kaboom was probably 100 bar or more. Beautiful, but for Starship definitely a non-starter. Over at NSF people are discussing launch abort systems for Starship, which has always been an issue because despite the vehicle's excellent payload, the nose section is SO large that a functioning abort system would really cut into your mass budget. And then someone came up with what has to be the best solution I have heard yet. That's.......tremendous. Rubber or plasticized petrol for the fuel, the existing LOX header tank for the oxidizer, pyro torches for instant ignition. Hybrids have remarkably good TWR and can use differential throttling, vectored nozzles, or both to provide pointing. Detcord just above the top of the methane tank for separation. It separates with its forward flaps, which assist in maintaining a heading until escape engine shutoff. Multiple large chutes deploy from explosively-jettisoned panels under the LOX head tank and the aft of the fairing provides several meters of crumple zone for a survivable, if unpleasant, landing. Not good enough for P2P, of course.......but absolutely good enough for a 10-person crew.

I think the spreading of the aft flaps helped with stability actually. Here's how a three-piston ITS-style solution might work.

I think this is the leeward view. The heat shield would cover the leg completely. In three dimensions, there might be room to do something interesting. Two pistons that descend vertically and one that protrudes through the skirt (perhaps from the thrust puck) to the furthest extent of the foot.

Some good notional renders from Twitter: The fold-out legs a la Falcon 9 would be the simplest, sure, but they're also simply not possible because that would involve (a) a seam in the heat shield, and (b) the entire vehicle resting on the freaking heat shield tiles. His pop-out legs a la New Shepard are also a nonstarter because of the heat shield seam issue and because of load paths. So the straight-down solution seems straightforward. It's also the most capable of auto-leveling and shock absorption. ITS-style legs that descend straight down and then fold outward are another possibility.

They're the same shape on the lee side and the windward side, so presumably they have a common design. The only way for the windward version to get out of the shielded fairing would be to extend downward first. So clearly they cannot be hinged at the skirt as with the F9 legs. So they may extend down and then out like the ITS legs. Or, they may extend down and then out, similar to the legs on New Shepard.

Here's a gif of the ITS landing leg deployment if you wanna compare.

The RL10C-1-1 is 2/3 of the dry weight of the RL10B-2. It's also substantially smaller: 59% the length and 73% the width. The nozzle is fixed rather than extensible, which makes it simpler and reduces failure modes. The Centaur family of upper stages has always had very good dry mass ratios, which means reducing weight is super important.

ITS was going to be 12 meters, carbon-fiber, with no wings, and primarily intended for Martian colonization. Of course they had no budget for this and so they scaled back and built Starship. Originally designed to be carbon-fiber with PICA-X heat shielding as well, they switched to stainless steel and ceramic tiles because it works better and is a lot cheaper and easier to build. They added the flaperons to control attitude. Unlike ITS, Starship has valid use cases for LEO and cislunar activities. Here's the image from the 2016 IAC showing those giant landing legs:

If they are more like the old ITS legs:

More evidence that the engines did not contact the ground at touchdown: If they had, there would be vertical crumple damage around the mouth of the engine bell.

I'm not saying it was properly choked or anything. But yes, if you look at slow-motion captures it definitely appears the bulk of the explosion happened in the LOX tank.

I have said this a thousand times but I will never understand why they didn't just build Jupiter DIRECT. Keep building virtually identical SRBs and SLWTs, design a thrust puck adapter and a payload adapter, and mount the early Orion CEV on top using the exact same Shuttle OMS engines to power Orion's service module and complete orbital insertion. Exactly the same launch profile as the Shuttle. Exactly the same contractors as the Shuttle. We would have been flying crew from US soil to the ISS this whole time and building experience toward something like the SLS but far more capable. And we would have had time to upgrade Orion's service module, too. Nope, no one does because they haven't even picked a design yet. The use of newly-built F-1A engines (two per booster) would work remarkably well and would largely obviate the inherent stupidity of using hydrolox sustainer engines. But that wouldn't let enough pork flow so they will probably try to go with advanced solids.