SpaceX Starship / Super Heavy Discussion Thread

SpaceX had a presentation last night about the latest news on their Starship/Super Heavy reusable rocket.

To get everyone up to speed:
Three years ago, SpaceX announced their plans for an extremely heavy launch vehicle that could put 300 tons into low Earth orbit reusably. Since then they have scaled down their plans slightly, due to budget and timeframe constraints, but they have converged on a design that should put 100-150 tons into orbit reusably. It consists of a two stage vehicle, larger than the vaunted Saturn V, with the first stage landing similarly to SpaceX’s current boosters and the second using a new system with a kind of aerobrake maneuvering system.

It’s difficult to overstate the importance to space flight if they can pull this off. Their goal is for the craft to be nearly as reusable as a commercial aircraft, with the costs approaching a small factor of the fuel/propellant costs. Despite the large size, the propellant should cost under a million bucks for a full flight. That’s only $10/kg. Flight costs of $100/kg do not seem unreasonable, which puts the system at a tenth the cost of SpaceX’s own Falcon Heavy system, which was already the cheapest around.

A very significant design change along the way was the switch to stainless steel for the main structure. Despite seeming like an old and obsolete material, stainless has a few key advantages here. First, it actually becomes stronger than even carbon fiber at cryogenic temperatures. Second, it stays strong even at high temperature–such as those of orbital reentry. It remains structural even when red hot. And finally, it’s cheap. So cheap that SpaceX can build a bunch of prototypes and not worry if they blow up or otherwise fail.

About a month ago, SpaceX performed their Starhopper test. It’s a simple test article that went up about 500 feet, demonstrating their stainless steel construction technique, their new Raptor engine, and their control systems. The test was a total success.

Their next test will be a full sized version of the upper stage, called Starship. They’re aiming for a 20 km height, which will allow some testing of their reentry systems. They’ll probably go higher than that eventually, but to get to orbit they need the first stage (the Super Heavy booster).

Speaking of reentry: you may notice from the picture above that Starship has four wings. Each of these wings (really, aerobrakes, since they aren’t airfoils) can fold at the base using powerful motors. The craft will steer by folding the aerobrakes in and out, a bit like a skydiver tucking his arms and legs in to orient. It does this to maintain a controllable descent through the atmosphere. Once it nearly reaches the ground, it does a flip and lands on its tail under rocket power.

You may also notice that the ship is very shiny, which of course comes from its stainless steel construction. Give it a very classic sci-fi vibe.

SpaceX still hasn’t said too much about their Mars plans aside from the launch system. They’ll need to work on that, but I hope that once they get the thing flying, NASA will have no choice but to adopt it for their plans, and they’ll be able to work together on it. We’ll see.

And for anyone not interested in Mars at all, or even space, this craft will have one very tangible near-term benefit: the launch of SpaceX’s Starlink internet service. This requires *thousands *of satellites in orbit, and although the first batches are being launched on the Falcon 9, it pretty much requires Starship to complete the constellation. Starlink will enable gigabit internet for virtually everyone on Earth, no matter how remote.

Fantastic — thanks for sharing. It also prompted me to finally learn a few things about Musk — his life and accomplishments.

Dr. Strangelove - thanks for sharing this. From the GQ thread, I also read up on the stoichiometric ratios and have a better understanding of the turbo cooler configuration of lace. Sounds fascinating. I would really enjoy doing the controls design on the fuel side :slight_smile:

The L/D ratio of the rocket seems to be Lower than usual rockets. Is that correct and why is that ?

Also I have good experience with Stainless steel in all kinds of environments and agree with your post. We typically use SS316L for cryogenic applications. Is this also SS316L or some other grade - from the shine of it, it looks like a higher grade.

Also wanted to point out that from the video, you can see when the rocket changes from hydrogen fuel to some other fuel (maybe solid propellant) when the color of the flame changes from blue/almost invisible to orange.

I assume you’re talking length/diameter here, and not lift/drag.

There’s a pretty simple reason for this: assuming a roughly cylindrical rocket, an engine on the bottom has to lift the entire column of fuel above it. Because there are limits to engine performance, that means there’s a natural limit to the height of the rocket.

The Raptor engines that SpaceX is using are quite advanced, and so the Starship/Super Heavy combo will beat out the Saturn V a bit in height, even with the latter being somewhat tapered. The fuel mix is also different, though I don’t know how this plays out. Nevertheless, physics and current technology don’t let it be, say, twice as tall as a Saturn V.

So when you’ve capped out your height, you have to go wide, and Starship is not a slender rocket at 9 meters. Saturn V was 10 m at its base, so it was chubbier yet, but the Falcon 9 is about 2/3 the height and 2/5 the width.

Musk has said that the next-gen rocket will likely be 18 m. It was left unsaid, but the height will probably be the same. So it will be quite fat indeed, and about 4x the mass.

I think it’s a pretty standard 301, with some parts cryo-treated.

Musk said that the current Starship prototype has a 200 ton dry mass, but that they’re aiming for 120 tons or lower. That’s a big difference, but I suspect they’re not currently being too careful with the sheet thickness, and just rounded up to whatever they could get cheaply from the supplier. It doesn’t matter in the short term, but to save weight in the long term will probably require dialing in the exact thickness they need as a function of height (lower sections need to be stronger since they hold the weight of everything above it).

It’s all methane and liquid oxygen, but they may have varied the mixture ratio as a test. The Raptor runs at 3.8:1 nominally, which is a little rich: 2O[sub]2[/sub] + CH[sub]4[/sub] = 2H[sub]2[/sub]O + CO[sub]2[/sub] should equate to 4:1, so the 3.8:1 leaves some unburnt methane. Alternatively, it may have just entrained some dust while near the ground and caused it to glow orange.

Going along with SpaceX’s drive to minimize cost, methane is basically the cheapest hydrocarbon fuel you can get. Bulk prices are 20 cents/kg or less. In comparison, hydrogen is dollars per kg, and I suspect that solid fuels are tens of dollars.

Normally, all these costs would be in the noise. But with long-term reusability, propellant costs start to matter.

That and who wouldn’t appreciate being the first to launch the Shiny Winged Space Dick.

Aside from going to Mars, are there use cases where it’s advantageous to lift 100-150 tons in one go, aside from lower $/kg? Are there planned or semi-science-fiction installations which weigh 100 tons?

With the planned level of reusability, where will variable costs mainly come from? Does a lot of it come from replacing not-so-reusable parts? Having people test the systems before a launch?

I’d love to see a 100-ton space telescope. And of course things just generally get easier when you have to split them into fewer pieces. The ISS would be more interesting if it had some 100-ton modules; that’s likely big enough for a human-scale mesogravity simulator. That is, something that simulates lunar or Martian gravity. That might even be a prerequisite before sending loads of people to Mars.

Hard for me to say. I expect that there’s a really long tail of things to fix here. I think there are small armies of people doing all sorts of things, and each one will have to be improved and automated to really get the costs to the lowest levels.

Payload integration is going to be a big one. Satellites tend to have all sorts of unusual ways of attaching to the rocket, and these differences cost money. But maybe SpaceX could build a really flexible payload adapter; one that is suited for almost all payloads. It would necessarily weigh more than custom adapters, but with 100 tons, who cares? Design it so you can bolt almost anything to it, and give it enough damping so that even delicate payloads survive.

Containerization for cargo ships dramatically reduced the cost of transport, in part because there was just so much less manpower involved in handling everything individually. Space is going to be the same way, but it remains to be seen exactly what form that’ll take.

I can recommend the biography of Musk by Ashlee Vance, if you like that sort of thing. There’s a lot of good detail about the early days at Paypal, Tesla, and SpaceX.

I bet the Pentagon, CIA and NSA would too. From what I’ve gathered, telescopes tend to have no larger than 2.4m-wide lenses but you could have extremely high resolution optical sats if size/weight aren’t at a premium.

Use labor do something custom-made then formalize it so a machine can save labor by doing it instead of people. Kinda like Oxygen Not Included

Yeah, it made me think of containers too. Perhaps when launch cost are 1 order of magnitude cheaper if your sats fits the spacecontainer rather than the other way around, satellite designers will orient their design to making it compatible with a containerization-like arrangement. A kind of ISO/NATO-like standard which may not be optimal in any one case but standardizes an industry for smooth operations/low transaction cost interactions and ends up being worthwhile overall because it allows everyone to focus on taking known compatible inputs and turning them
equally well known and compatible outputs. You see something similar with financial instruments; Once they’re generic/compatible/known quantity enough to be traded without having to take a long amount of time, effort and thought to closely examine them, you can integrate them into large scale systems.

Thanks for that Dr.S, always a font of knowledge on such things.

Enlighten me, is stainless widely used outside of SpaceX’s vehicle? The benefits you mention make it sound like perfect choice but lovely shiney rockets have not been the default as far as I can tell so have I just missed them? or have it’s limitations been more relevant for other manufacturers concepts?

As far as I understand it, stainless is not an ideal choice if reusability is not a requirement. If you don’t care if your booster survives reentry and you’re using solid rocket motors, the extra weight of stainless provides no benefit over aluminum lithium or CF.

Here’s a link from everydayastronaut with more information:

It’s not super common. The Atlas rocket from the 60s used stainless steel, and indeed it was a very lightweight design–in fact, it was nearly an SSTO! It had only a single set of propellant tanks, with the only staging hardware being a set of booster engines that were ejected in flight.

It had what are called “balloon tanks”; called that because almost all of their rigidity came from internal pressure. Starship is more conventional in that respect, as it has internal bracing and other structures. It still needs to be pressurized for flight, but it is strong enough not to just crumple when on the ground.

As YamatoTwinkie says, some of the advantages to SS are exclusive to reusable flight. In particular, the strength at high temperature. Reentry is tough because you’re trying to shed so much kinetic energy. One approach is to just insulate yourself–Starship will be using that in part, with insulating tiles. But the parts that are bare stainless will just get hot and radiate the energy away.

Radiative cooling gets much more effective as the temperature goes up. At 300 C, you radiate 16x the energy than at room temperature. At 600 C it’s 81x, and at 900 C it’s 256x. 900 C is probably too hot but 600 C may be doable.

At the other end of the spectrum, stainless gets stronger at cryogenic temperatures, to the point where it matches carbon fiber at a certain point. This is of course only useful if your propellants are cryogenic (or at least one of them), which is the case here (liquid oxygen and methane).

It does seem like the cryo benefits will lessen as soon as the rocket launches, since the propellant levels will go down and the top parts of the tank will have to heat up. I’m curious how SpaceX is modeling this. They probably only get the maximum benefit on the lowest part of the tanks since these will have some propellant for virtually the whole flight.

Yep. It’s also exactly how cubesats work, so it’s not entirely theoretical. There’s a surprisingly short spec document that gives all the limits on dimension, mass, materials, etc. and you simply design to that. It’s not as mass-efficient as a custom design would be, but it doesn’t matter. It’s good enough and the cost savings enables just about anyone to build a small satellite. There’s a whole cottage industry around the spec, which means you can use off the shelf parts for anything you don’t want to build yourself.

With current launch prices, that argument doesn’t quite extend to satellites that are a meter or more on a side. Maybe with Starship, though.

All good stuff. The natural assumption for advanced space vehicle construction would be for exotic composites, alloys and ceramics. To find out that it is good old stainless steel is somehow comforting, not sure why.

My nephew works for SpaceX, and he’s in several of the photos in this article. Really nice photography of the spaceship.

Well, there’s at least one good practical reason for that–because stainless is such a common material, it’s very well understood and there are lots of people that can work with it. SpaceX put together Starhopper and the Starship prototype in an open field, with concrete and wooden jigs, welding bits together with TIG or MIG, etc., by people more used to welding together water towers than aerospace equipment.

Aluminum-lithium, in comparison, needs exotic methods–SpaceX uses friction stir welding, which requires a giant robot arm and a bunch of other advanced equipment. It obviously works, but it has a huge capital cost, with only a small community of people that know how to run that stuff, and probably substantially more stringent quality/inspection requirements.

It’s not at all unreasonable to be comforted by materials that are so common and well characterized, as opposed to ones that are so exotic that hardly anyone knows what can go wrong with them, let alone actually how to work with them.

Awesome! Tell him to keep up the good work. It’s all very exciting. SpaceX is one of the few organizations that really make me optimistic about the future of space flight.

The SpaceX website has a nice page on Starship/Super Heavy now.

I almost forgot–the Centaur upper stage also uses stainless. It’s a great stage, still in use and likely to still be in use years from now. Well, it’s great aside from the cost, which is astronomical. Er, stratospheric? Sky-high.

Speaking of cost, here’s one crazy claim:

Even hitting the $1M/Raptor target is nuts. That’s roughly what their existing Merlin engines cost, but Raptor is about twice the thrust, significantly more efficient, and vastly more reusable.

It’s also vastly cheaper than any of the competition. For example, the Russian RD-180 engine, used in the Atlas V, costs at least $10M but is under 2.5x the thrust. Which makes the Raptors 1/4 the relative cost, for an engine that’s far more advanced. If they can hit their “$1000/ton” target, they’ll be 23x as cost efficient as an RD-180. And the RD-180 is relatively cheap as far as competitors go.