sevenperforce

Okay, this is about as good as it gets for now.

Does anyone know what the probability curve for this sort of re-entry prediction is? Like, is it "basically anywhere in this window with equal probability"? Or is it a bell curve?

Yeah, with the exception of like one pass over Spain, Europe is in the clear. Actually I may have misspoken in that tweet. The tightest cluster would be apogee, not perigee, since it is moving fastest at apogee (the dots on the lines are five-minute increments).

Narrowed to a 42 hour window. It also looks like they have narrowed to 29 total possible ground tracks.

I disagree, when you think of it, starship is a slow moving object that's basically falling through the air column and it can only steer itself to a certain degree where falcon9 is a supersonic dart, its very skinny whereas starship is fat. You can give starship the best software and engineering in the world but it physically has a larger margin of error and that will only grow when it is going faster and higher. I agree with @RCgothicthat Starship has less of a wind problem on descent problem than F9 but I don't think fineness ratio is the operative issue. Fineness ratio is an important variable on ascent because a skinnier rocket has a larger bending moment relative to high-altitude winds and so structural stability (not guidance) is an issue. On descent, structural stability is no longer the issue; it's a question of guidance. Starship falls more slowly than F9, yes, but it is also much heavier with a more even distribution of mass. A supersonic dart with all the mass on the prograde end is extremely susceptible to wind buffeting its "lightweight" end and causing torque around its center of mass. In contrast, even very high winds aren't going to change Starship's inertia very much. Eh, the entry burn is just a constraint imposed by the relatively low strength and temperature resistance of aluminum-lithium alloy. I'm still amazed that those little legs caught it successfully. Although the two-engine landing burn probably helped with that:

Agreed. They will certainly pull these engines off and reuse them, assuming there is no damage. The Raptors are expensive. But the rest of the vehicle is fairly disposable at this point. They need data now.

I watched the launch and landing on a TikTok live and it was amazing! Go SpaceX!

And there's the tri-vent.

Tri-vent should have happened by now so it looks like a hold.

Yes, I've always lived in America and it makes no sense to me for the "first floor" to be anything other than the first floor you can enter, the "second floor" being the second floor you can reach, and so forth. The Mercury capsule had posigrade separation motors but I don't think they provided much total impulse relative to the booster. That booster did end up in orbit, but it only would have hung out in space for a few orbits because they deliberately targeted a low enough perigee to ensure that John Glenn made it back to earth promptly if the retrograde deorbit motors failed. In contrast, the payloads for Sputnik 1-3 were not intended to be recovered and burned up on re-entry, so they didn't care where they came down. Thus they put them quite high, and the boosters stayed in orbit for a while. Most people who reported seeing Sputnik 1 were actually seeing the R-7 core (which was actually about the same length as the Long March 5B core but had a mass less than 8 tonnes and an engine cluster 83% lighter). Sputnik 1's R-7 core stayed in orbit for two months but re-entered without incident.

Russians count the stages in the order they fall off: one, two, three. Americans count the stages in order of how close they are to the ground: boosters, one, two. So we say that the SLS and the Delta IV Heavy and the Falcon Heavy and the Soyuz-FG are two-stage rockets with parallel boosters on the first stage. Russians say all those are three-stage rockets. What I mean above is that the Long March 5B configuration (what Americans call a single-stage rocket with parallel boosters and what the Russians would call a two-stage rocket) is the first rocket since 1958 to have its central core fire all the way from the ground and end up in orbit. The R-7 Semyorka, which launched Sputnik, Laika, and Sputnik 3, had the same configuration as the Long March 5B with the central core reaching orbit. This is generally regarded as a bad idea. The R-7 launched several times after that in its ICBM configuration but never went to orbit.

Unless I miss my guess, the Long March 5B configuration is the first launch vehicle designed to orbit its entire first-stage core since Sputnik 3 in 1958.

He knows this.

Looks like drag is pulling the stage down a little faster than expected, narrowing the impact window from about three and a half days down to just two days. It should re-enter at around midnight EST on May 9, give or take 24 hours. Spain, Portugal, Turkey, and Italy's boot are the only parts of Europe that are in the zone. All of Africa, Australia, the Middle East, most of China, and Southeast Asia are in the zone. Mexico, most of South America, and all of the United States south of Michigan are in the zone.

Engine-out is unrecoverable for Falcon 9. Engine-out is fine for Starship as long as it happens at relight (not mid-landing-burn) and the failed engine doesn't frag the others.

The low-slung effect seemed REALLY promising to me. A bare-bones lander with egress at the surface could enable transfer to a pre-placed hab via pressurized rover. Just imagine -- Starship reusably delivers a big surface hab with its own airlock and a couple of semi-autonomous solar-powered pressurized rovers with docking ports in the tail. Orion meets the Dynetics lander at Gateway and you hop in. Basically it's just a lander can, nothing more, with two or three sets of drop tanks. It only needs consumables for the trip down and the trip up. It takes you down to the landing site (dropping tanks along the way) and then once the dust settles the pressurized rover pulls up, attaches to the front of your lander can, and you climb in and drive over to the hab which is a safe distance away. At the end of the mission you can climb back in the rover and repeat the process in reverse. Plus, you get back to the Gateway and it can replace your tanks with a Canadarm for the next mission. Maximum simple reusability, minimal complexity, no wasted mass being dragged up and down. There really is a dearth of good engines in that particular space, though. Like, a real lack. We don't have many options for hydrolox engines at all...until BE-3U flies, the only American upper stage hydrolox engines we have are RL-10 variants, I think. Vinci looks very promising even if that long combustion chamber is scary.

I can't remember if I've done the math on this before or not.... 120 tonnes dry mass, 30 tonnes landing reserve prop, and 100 tonnes payload to LEO, so mf is 250 tonnes. 1200 tonnes prop at staging but that presumably includes those 30 tonnes in the header tanks so m0 is 1420 tonnes. So dV expended to LEO is 6473 m/s. We know that Starship needs all six engines to light at staging to help avoid overmuch gravity drag, which means a fairly lofted trajectory and some gravity drag losses -- probably as high as 600 m/s. So we can ballpark staging velocity at LEO - 5873 or roughly 2 km/s. Superheavy needs to cancel its downrange velocity (but not its upward velocity) and head back to the pad. A lofted trajectory helps with this because it only needs enough speed to traverse the same distance during its hang time. We can assume that Superheavy, like Falcon 9, will have a landing burn that kicks off just after the vehicle goes subsonic, around 310 m/s, but just as Falcon 9 it will need about 200 m/s more to account for gravity drag during the landing burn. In the NROL-108 mission, Falcon 9 needed about 550 m/s for the landing burn (it was a single-engine burn) and also burned all three engines for an entry burn for 28 seconds. It is 27 tonnes dry so that is 6 tonnes of prop for the landing burn, and a Merlin 1D has a mass flow rate of 305.4 kg/s so burning three of those for 28 seconds was about 25 tonnes, so total mass before the start of the boostback burn would have been 58 tonnes. Its boostback burn lasted 37 seconds and thus would have consumed 33.9 tonnes of propellant. Since the vacuum specific impulse of the SL Merlin 1D is 311 seconds, the boostback burn provided about 1400 m/s (staging velocity was 1600 m/s but part of that was an upward component, because lofted trajectory). The SL Raptors on Starship light up at 330 seconds and reach somewhere around 356 s by burnout. It won't need an entry burn. I'm estimating Superheavy at around 290 tonnes dry and 3690 tonnes wet, so it burns around 1700 tonnes (49% of its prop load) getting to Mach 2, another 1450 tonnes (43% of its prop load) getting to staging, about 229 tonnes (6.7% of its prop load) on the boostback, and about 49 tonnes (1.5% of its prop load) for the landing burn.

It will impact between 4 AM on May 8 and 2 PM on May 10, EST. It's a large window because it's tumbling but those are the upper and lower bounds based on observed decay rate. I feel like a three-tonne engine might get melty during re-entry aero but I can't imagine it melting enough to actually come apart.

Something else to keep in mind -- those six RL-10s on the top of the Saturn I were RL10As without nozzle extensions (you can see if you look closely) so they were completely regeneratively-cooled, allowing them to be clustered like that.

Expander-cycle engines are great for high-energy orbits. They are lightweight and ridiculously efficient and just absolutely amazing pieces of engineering. Even if they are expensive. They just really, really suck for medium-energy high-mass payloads.

We don't know whether the engines will separate from the thrust structure or not. We don't know when in the breakup sequence the stage will break apart. We don't know which part of the hot streaky bits flying through the upper atmosphere at Mach 20 are the engines. So yeah, there's no way we can hit the engines exactly. My son's birthday is that day and I'll be with him in Virginia so I hope we get to watch it zip over the Virginia skies in daylight and dump itself safely in the Atlantic.

But remember, SLS is sooooo safe. It's so safe, in fact, that it cannot pull off a free-return trajectory without sacrificing so much of Orion's propellant that Orion can no longer enter lunar orbit.

The question is not whether we can hit it -- I am pretty sure we can hit it because it's pretty big and coming in slow compared to ballistic missiles -- but whether we can hit the engine cluster exactly. For a nuclear ICBM, the only concern is disabling the warhead. That's why the Nike Sprint and other similar ABMs carried nuclear warheads of their own -- even at quite a distance, a nuclear blast would produce enough trauma and prompt x-rays to the incoming warhead that it would no longer be able to detonate. The risk of the Long March 5B core stage is an impact to civilian areas by the engines. They are the only thing we know about that is definitely going to frag anything they hit. Unfortunately the only way to blast them into little pieces is to score a DIRECT hit with a high-energy warhead. Hitting any other part of the stage might change the trajectory of the engines but it won't appreciably damage them. Remember, this is what one of the powerheads from one of Shuttle Columbia's RS-25s looked like when they dug it out of the crater it made: That's just the preburner and turbopump; the rest of the engine disintegrated on impact. Hence the (substantial) crater. You'd need a direct hit from a shaped-charge warhead to actually break apart an RS-25 preburner. The twin YF-77 engines on the Long March 5B core stage are only slightly smaller than an RS-25. Similarly, here's the impact site from some of Columbia's OMS propellant tanks: I don't think the Long March 5B core has any hypergolics, so that's good. It does have at least one helium COPV, though, which will ruin anyone's day if it hits them.

I said specific impulse above when I meant to say mass but otherwise it was a decent explanation

That's the point, I think. Even if Starship literally NEVER launches from Earth with crew, it will still be completely transformative.