And they plan to catch the pointy bit, too.
Yep. Been the plan from the beginning.
However, I am very curious about the approach they’ll take. The booster is mostly cylindrical, with the grid fins at the top being the only parts that stick out significantly. So they could easily install a couple of load points below those that the booster is caught by. Also, they can be fixed in place, since reentry is a relatively gentle process. It doesn’t need a special heat shield or anything.
But the Ship has a delicate heat shield covering more than half of it, and four flaps that stick out significantly, and anything that sticks out has to be shielded vs. mach 25 reentry conditions. Some have speculated that they’ll catch it on the flaps, but they seem too delicate to me. Plus the arms have to close in such that they don’t catch on the lower flaps first.
Maybe there will be some load points that mechanically extend out once it’s through reentry, and for the last step it’ll hover long enough for the arms to swing in between the top and bottom flaps… but it’s still hard to imagine how it all works out. Maybe they’ll have some completely different system. No idea yet.
Some interesting bits that came up recently. SpaceX claimed in a court filing that they’re spending around $4M a day on their Starship program:
So, roughly $1.5B per year. We’ve heard rough estimates before about total program costs of $5-10B, but that would be over several years and we haven’t heard an annual number until now. Regardless, that’s pretty cheap for a development program like this.
And then there’s this, with regard to the SLS rocket:
A tower–probably less complex than the Starship tower, since it doesn’t have any “Mechazilla arms” or anything–is now estimated to cost $2.7B by itself. Absolutely insane.
SpaceX’s tower could not have cost more than a couple hundred million, since they built a second one in a matter of months earlier in the year. Given all the other stuff they do down at Starbase, the tower could only have taken up a relatively small proportion.
There seems to be no limit to the inefficiency of traditional government contractors like Bechtel.
I just checked, and my guess was basically on the money. Went through some video feeds and just after the 3-engine burn starts, I see a deceleration of ~17 km/h/s, or 4.7 m/s^2. Just under half a gee.
That’s about the acceleration of a car with a 0-60 time of 5.7 seconds. Though what you’d feel is substantially more, since you have to add gravity. The engines are putting out 1.5 gees thrust.
An observation about catching both the Super Heavy booster stage and the Starship payload stage …
The two halves of a single launch will arrive at different times, perhaps separated by days; certainly by an hour. But I’m struggling to imagine recovering the payload onto a launch platform that already has a booster sitting on it. Whether said booster is freshly landed and awaiting post-flight inspection / refurb, has its refurb already in progress, or is itself launch ready once the rest of the stack is attached on top of it.
So I’m left imagining two platforms, one used to recover used boosters and also to launch fresh stacks from, and a second different sort of platform / tower just to recover returning payloads.
Which of course means the crazy inefficiency of then craning the payload off the recovery platform, doing whatever inspection / refurb process, then restacking it by crane atop a booster on a launch platform.
All of which sounds positively Old Space in its slow tiresome process complexity.
What am I missing?
If they really:want to impress they could land the Starship on top of the booster.
“Hot re-stacking.” The best part is no part.
Bechtel has a huge non-governmental construction business as well–the energy field comes to mind. The private sector would not be buying construction from Bechtel if it was much more inefficient than other contractors.
So the problem is not Bechtel, the problem is NASA, federal government procedures, Congress…
You’ll need to toss some kind of interstage in there, at least with the current hot-staging design.
Or Bechtel’s incentives, at least. They may behave differently with NASA because they have no interest in doing otherwise. A good contractor will seek to understand the client’s limits and push back on dumb requirements if it’ll blow through the budget. But here, the client has an unlimited budget and has promised to pay all of their expenses, plus a profit, as long as they can be properly documented. So Bechtel has no reason to ever tell NASA that what they’re asking for is dumb and will double/quadruple/whatever the cost.
People on various sites have been asking for the “buoy cam” footage enough that it’s become almost a meme. And we finally got it!
Quality isn’t great (probably because the source footage was a 360 degree cam), but you can still clearly see it start in bellyflop orientation, then do the backflip, and finally soft-land in the water, with some of parts of the vehicle still orange-hot from reentry.
Nevertheless, I think that’s the idea. The catch arms can swing left and right quite a bit. So they catch the booster, then swing it over to place on the launch mount. That takes well under an hour. They swing back in the other direction, catch the Ship, and place it back on top of the booster.
That sounds crazy, but the booster already demonstrated accuracy of <10 cm. So why can’t they just catch the Ship >20 meters away? The engine thrust will largely be vectored to the tower, not the booster beside it, so that’s not a problem.
But maybe that’s still too risky, at least for the first several flights. The alternative is that they just catch on the second tower, which isn’t too far away (but far enough that a crash won’t cause too much damage). They use the catch arms to lower the ship down onto a ground transporter (an SPMT), drive it over to the first tower, and then use the catch arms to raise it back up again. Would add some time to the process, but probably only on the order of an hour. They’ve demonstrated for a while now that they can shuttle boosters/ships around quickly using this method.
All of that still leaves the question of inspection. At the moment, it seems to me that they have a clear path to a booster that doesn’t require more than a few self-checks. The damage we saw was minimal and superficial and it seems clear that they’ll figure out solutions easily. The landing burn was perfect (all 13 engines lit) despite the orange-hot reentry. The outer ring of engines were damaged but that can be solved (I wonder if they’ll add gimbaling hardware, so that they can point exactly in the direction of the entry forces). The Raptor 3 engines will help, too.
The Ship is a different story. It’s still experiencing flap burn-through, and missing tiles, etc. They’ll want to inspect these and repair them. But it also seems like they can just mount some good cameras and maybe some LIDAR units to the catch arms to perform an automated inspection, and then go in and do what needs to be done. We’re still talking hours, not days or months.
Oh, and finally there’s the question of the payload. For tanker flights there’s nothing to be done: their payload is the remaining propellant, which gets transferred over to the depot in orbit. The next major payload are Starlink satellites. They plan on having a “Pez dispenser” system, but I still don’t see them loading that on the launch pad. So probably they’ll wheel the Ship over to a loading bay and insert the satellites there. If they wanted to be super rapid, they could have one Ship loaded and ready to fly as soon as the booster comes back. The returning Ship gets brought over to the bay once it returns. But that might be overkill.
Other satellites, etc. will undoubtedly have to be integrated in a dedicated building. Some require delicate handling, a relatively clean environment, etc. No way around that being a fairly slow process. But once ready, the loaded Ship can just take the next available booster.
That’s still a problem at NASA, though. They should be structuring their contracts so the contractors have an incentive to deliver. If they’re not, that’s on them.
Will the tankers be a dedicated version optimized for carrying fuel, like stretched tanks for carrying the payload fraction as fuel? Or will it simply be a standard Starship omitting payload and thus arriving on orbit with a fuel surplus? For the latter, would the reduced gross weight make it worth it to omit an engine or two, saving more weight and fuel?
Mostly. But SpaceX has always demanded fixed-price contracts even when their competitors suckled at the cost-plus teat. Bechtel could have done the same. They have enough experience with private contracts that they should be able to submit an accurate fixed-price bid. It’s still true of course that NASA (actually Congress) holds the ultimate power, but the contractors still bear some responsibility for going along with the system.
There’s supposed to be an optimized version eventually, but I think at first it’ll just be normal versions that just arrive with a surplus. Musk at one point said a dedicated tanker would look really weird, but didn’t elaborate.
I’d guess they aren’t going to bother with any half-measures like stretching or reduced engine counts. Too much trouble for not enough benefit. If they’re going to do it, then do all the things at once.
Yeah, upon reflection SpaceX seems to have mostly avoided the “squeeze out another kilo of payload” mantra that characterized the adapted-ICBMs of Old Space. The proponents of “Big Dumb Boosters” for decades tried to tell anyone who’d listen that the cheapest way to increase payload was start over with a bigger rocket, but no one listened.
Great explanation; thank you!
Aaah hah! The rotation feature was totally missing in my mental model. I figured the chopsticks were fixed over the flame diverter and that was that. So the booster would be caught a few feet above where it would be set down to be refueled and relaunched.
Now that I understand they’re catching on a different azimuth and carrying the caught unit, booster or ship, around the tower to restack it, the whole plan makes complete sense.
It’s a bold plan, Cotton; lets see how it works out for them. Seriously, I have no doubt they’ll make it work. Maybe even the first time.
[aside]
Today I was following a Model S down the street. It had the usual Tesla stylized T logo low on the trunk / hatchback lid. And the usual “Dual Motor” label near the right tail light.
Over on the left where it usually says “Model S” or whatever there was a SpaceX logo in the same raised chrome style and finish. Looked like it belonged there. Very cool.
All you Tesla-driving space nuts need to get one. I would.
You can sorta see what’s going on with these two pics:
In the first, you can see that the arms (which are directly over the launch mount) are somewhat “clockwise” of a pure radial-out position.
And then in the second, you can see they swing quite a bit counter-clockwise, to the point that they’re well clear of anything on the mount.
It’s definitely still a bold plan, and as I said, they might decide to land on their second tower, at least for the first several tries. But the articulation they’ve designed in at least allows for it. And they can still use the arms to pick stuff up from the ground and set it on the mount, or vice versa. No extra crane required.
The arms themselves have screw drives so that the craft can be brought in or out (or rotated) for adjustment.
Neat. The two companies actually share tech, so it isn’t entirely unjustified. The grid fins on Starship are powered by Tesla drive motors, for instance.
Another pic showing just how far the arms can go:
In that pic, they were picking the booster off the ground to place it on the mount. But they should be able to catch it in that position as well.
Well, possibly the booster will have to be caught in a more radial position due to the way the forces are transmitted to the tower. But the Ship weighs less.
Cool. Thanks.
Do not leave the smart people alone with a few billion dollars or maybe, just maybe, something spectacular will happen. Especially when top management isn’t trying to simply defraud the investors out of that money.