sevenperforce

Some of the most beautiful (non-official) renders I've seen so far. Full posting: https://www.humanmars.net/2019/10/renders-of-spacex-new-starship-design.html

As @5thHorseman said, you can't use quantum entanglement for communication, because even though the particles "communicate" with each other in a probabilistic sense, they cannot be used to pass information to an observer. However, we don't really know how time dilation (whether from inertial acceleration or gravity) affects quantum entanglement. There's ongoing research to tackle exactly that.

Lower-stage failure in flight is pretty benign, as rocket accidents go. Plus, Starship is quite aerodynamic. The real challenge is zero-zero abort, where you not only need enough thrust to escape the fireball on the pad (think Amos-6 but in the lower stage), but you need to get far enough away from the affected zone to land safely. Three atmospheres, so yeah, it's not bad. The big risk for a non-hypergolic vehicle like Superheavy is that the fuel and oxidizer mix before ignition. That's where you get a bomb. Worst-case scenario for a pad emergency would be if the inner bulkhead ruptures without igniting and the LOX and CH4 start to mix slowly. Then you're stuck on top, without enough thrust to get free of the pad, waiting for three thousand metric tonnes of accidental high-explosive gel to detonate. The largest non-nuclear explosion in history, the July 3 1969 N-1 failure, occurred with only 15% of the RP-1 burning. If just half of the propellant in Superheavy formed an explosive gel and detonated, it would produce an explosion four and a half times greater than the N-1's, roughly a third the yield of the Little Boy bomb dropped on Hiroshima. This threat of mixing-then-detonating is one of the reasons why early rocket designers preferred hypergolics. Not only are hypergolics easier to design engines for, but because they ignite on contact they will never be able to mix-and-explode like cryos. The Titan II launch vehicle used for Gemini was fueled with hypergolics, which contributed to the designers being okay with ejector seats rather than a full-envelope LES as with the Atlas that launched John Glenn et al during Mercury. Counterpoint: if Falcon 9 or the Atlas V N22 violently turn to the side and snap in half, Dragon 2 and Starliner (respectively) both escape successfully with no LOC. Dumping the fuel would take too long. They would never put SRBs on there, but they might very well do six SL raptors instead of three. I could see multiple configurations based on use case: The tanker has no cargo so it doesn't need any 0/0 abort and can have a TWR less than 1 at sea level. The cargo starship at least needs a fighting chance and so it has an additional SL engine to allow it to slowly pull away. The cargo version would also be uniquely suited for massive deliveries to the lunar surface given its better hover capability on a single engine in low lunar gravity. The crew version not only has the heritage of the tripod engine configuration from the tanker (as well as Mk1/2 for that matter), giving it the highest data availability and proven reliability, but it would have three fixed, simplified SL engines that cannot throttle but are uprated to 2.45 MN. They would fire for only a few moments at nominal staging to really give Starship a spry jump away from Superheavy, allowing SL-engine shutdown to occur sooner and take more advantage of those vacuum engines for the rest of the ascent, and would give vital thrust in a 0/0 abort.

oh, ok (still kinda odd that it isn't mentioned on the mission diagram on the website, but I'm assuming someone ran the numbers) They left it out. You need 2,730 m/s out of LEO to get a lunar flyby. I don't know how much the crewed version of Starship will mass, but even if there was literally no crew cabin at all, an empty Starship can reach orbit with about 100 tonnes of propellant remaining. That only gives you 2,260 m/s. (380 s isp, 220 tonnes wet mass, 120 tonnes dry mass) And that's with no cabin, no passengers, no consumables, and no landing propellant. Unrelated: ripped the gif showing the Starship lunar landing with self-adjusting landing legs.

I'm not sure that it's "not in the current design" as much as it has simply not been made. They haven't built any vacuum Raptors yet. Here's where Elon explains... Current Starship design shown the other night is three RVacs. Burning those by themselves would be a T/W ratio of a measly 2.43 m/s2, which is simply not workable. A single-engine Centaur has a far higher T/W ratio and it is famously underpowered and needs an inefficient lofted trajectory.

While I agree with most of your rant (lol), they do need orbit refueling for #DearMoon. They just don't need elliptical orbital refueling.

This is also particularly likely if they put the LOX tank at the top rather than the bottom. I can't recall the exact configuration, but if the LOX tank is at the top then you have a massive barrier between you and anything explodey. Granted, LOX is awful, awful stuff, but there's no way to make it detonate under those circumstances. I would still imagine that having positive T/W is important so that you can actually separate from Superheavy. What if Superheavy catches on fire but remains standing?

Scott Manley suggested that and Elon said no, they would go dual-bell instead. No meaningful difference in efficiency between a dual-bell and a SL bell other than added dry mass. Also wouldn't work because They are regeneratively-cooled nozzle extensions, which means they are plumbed for pumping propellant through for cooling, which means you can't break them They are welded to the skirt of Starship

You only need a 64% fuel fraction to make LLO to surface and back on a single-stage with a vacuum Raptor. Descends to the ground at an impact of speed of thirty miles per hour directly into the flaming, exploding, shrapnel-shredding flaming carcass of Superheavy? More likely that they will add three additional engines for crew and/or fly them without as many props.

Basically reverse staging. I like it. I have done this exact same thing more times than I can count in KSP.

Elon has always insisted that if there is a problem with Superheavy on ascent, the upper stage can just boost way. But it cannot do that if it cannot pull a gee. If they bump up from six engines to nine, we could see a high enough TWR to be useful and potentially single-stage P2P. Shame no one talked to Elon about P2P on Saturday night. At least the legs will self-adjust.

The abort scenarios are starting to worry me. Even with all six Starship engines pushing 2 MN, we'd be looking at 12 MN divided by the stack's 1420 tonnes, which is less than a gee.

Not quite that speedy. On Saturday night Elon said Mk1 and Mk2 were very similar test articles for demonstrating skydive, turn-burn, and landing; Mk4 would likely be the first to orbit. He said they will build Mk1-4 before starting on a Superheavy at either site. I am guessing that Mk3 is the first to mate to Superheavy, but only if they get enough data from Mk1 and Mk2. We would potentially see propellant transfer demonstrations between Mk4 and Mk5...potentially 3 & 4 if they do end up sending Mk3 to orbit after Mk4. Mk1-2 are built with welded-panel construction; Mk3 and onward will be built from single-seam roll steel rings. The rings for Mk4 are already in place at Florida even though they are not yet done with Mk2. Elon said that because building with steel is so cheap, they're basically just going to keep building at max speed from now until they have a full fleet. Limiting factor is the Raptors. They are building a few Raptors a week now but need to be at a Raptor per day to get enough for a couple of Superheavy boosters. That's one reason why the build sites are going all the way up to Mk 4 before either starts on a Superheavy.

Animation on one of the sub-pages shows self-leveling landing feet descending vertically out of those sheaths for touchdown on lunar regolith: Evidently still planning on a chomper payload fairing rather than the cargo bay. It makes attachments easier, I suppose. 9 meters is only 2 meters larger than New Glenn's. Then again I am "only" just over 2 meters tall, so I guess that is a big difference. Length-wise, though, I think it is roughly the same as FH.

Good point. Titan is a much better bet. I'm wondering about the upcoming flight schedule. In my head, I was imagining that Mk1 will perform at least one or two hover tests to validate terminal descent variables before progressing to the 20 km hop, right? That just made sense. Gradatim felociter and so forth. Though perhaps not; gradatim felociter is the other egomaniac billionaire rocketman's motto. If someone went wrong during the Mk1 hover test, they'd have no data and an LOV, whereas if something went wrong during the landing after the 20 km hop, they'd have the skydiver-descent validation and mountains of useful data and an LOV. Seems like it is to their advantage to go straight from static fire to the full 20 km hop. If the touchdown fails and you lose the vehicle, nbd -- you've got the data from the skydive envelope and the turn-and-burn, so you update the landing info for Mk2 and iterate while you complete Mk3. If everything works as planned, then you test Mk1 and Mk2 to destruction on increasingly challenging recovery envelopes (and perhaps successively hotter re-entries) while preparing to do an orbital test with Mk4. Not sure where Mk3 fits if Mk1 is successful, then, except as perhaps a hypersonic L/D test article for substantially hotter entry, or as a lessons-learned artifact from whatever Mk1 and Mk2 are able to teach them. UPDATE: New SpaceX Starship website! Lots of goodies. https://www.spacex.com/starship

Diffuse is in the ballpark but not quite. Low density works, of course, but there is no monolexical antonym.

Usually, you cannot safely fire a vacuum engine at sea level, because the overexpansion of the flow will lead to it being "pinched" by atmospheric pressure and separating inside the nozzle. This flow separation is chaotic and so it can produce different forces on different parts of the engine bell. With the amount of force in a rocket engine, even a tiny bit of misalignment on the force vectors will easily rip the thin bell apart. Elon has suggested the possibility of using a dual-bell nozzle in the vacuum Raptors to allow them to fire safely at sea level. A dual-bell is a means of altitude compensation that puts a tiny convergence crimp part of the way up the bell, creating a defined surface where the flow separation can attach. It's very very slightly less efficient than a proper vacuum expansion bell, and of course the added weight of the outer bell portion is useless at liftoff, but it's a good tradeoff for a sustainer stage and is lighter than something like an aerospike. The bigger reason is raw thrust. The sea level Raptor has a ridiculously high thrust to area ratio for a booster-class bipropellant liquid rocket engine. It has a nozzle that opens 40% larger than the Merlin 1D but produces 130% more thrust. For example, the BE-4 has a nozzle that opens 46% larger than the Raptor's but produces only 25% more thrust (and may ultimately produce less thrust if Raptor's chamber pressure is upgraded). SL Raptor has a thrust-to-area ratio that is 135% greater than the mighty F-1. In fact, the SL Raptor's thrust to area ratio is a whopping 39% higher than that of the Space Shuttle SRBs themselves. But going from a sea level nozzle to a vacuum nozzle is a huge loss of this critical value. Recall that the Merlin 1D Vacuum engine bell takes up almost as much space as all nine lower-stage engines, yet produces only about 1/8 as much vacuum thrust as those nine would (the Merlin 1D Vacuum's nozzle extension gives it about 12% more thrust in vacuum than the underexpanded SL engines). The vacuum Raptor will only have about 8% more vacuum thrust than a SL Raptor. This is because the chamber pressure is so high that there aren't as significant expansion losses. Even though it's only an 8% higher vacuum thrust, a single vacuum engine would take up essentially the entire space occupied by the three sea level engines (plus substantially more vertical space) with no room to gimbal. It would also have no redundancy; those three gimbaling SL Raptors have enough thrust to land safely even with an engine out, and they have roll control, while a single vacuum raptor would not. This thrust is important for other reasons, too. The Starship needs a LOT of thrust at staging so the three SL Raptors will actually fire with the vacuum Raptors after Superheavy separation until their added thrust is no longer needed. Swapping the three SL raptors out for a single vacuum Raptor would lower thrust at staging by 31% while only providing a negligible drop in wet or dry mass, with noticeably higher gravity drag losses due to the loss of T/W ratio. Indeed. If reusing orbital stages was easy, then Centaur and Agena and Block D and Briz-M and DCSS and Fregat and Transtage would have all been redesigned for it. It was hard enough to find a way to land Falcon 9's first stage and that was the first reusable orbital-class liquid booster ever. Same dual-redundant spark torches used on Raptor, most likely. Alternately they could actually just rebuild the spark torches into the hot-gas RCS thrusters themselves. That's what those spark torches for Raptor are, anyway: spark-ignited hot-gas thrusters. I am certain that will be the method. I wonder if they would do aerocapture at Titan or try to do aerocapture at Saturn itself.

Probably no more jarring than chute opening but I could be wrong.

That only works if the hydrogen is already monatomic. Diatomic hydrogen is transparent to basically everything.

Right, down vector. You can point the exhaust nozzle down and you could also use a moveable panel to duct a driver fan down. Fan and rocket engine both either gimbaled or ducted down for liftoff, then realigned inline for horizontal flight.

That was hilarious.

Gotta seed the hydrogen with tungsten dust or something else that absorbs neutrons, but yeah. Using radiation for heating and separating the reactor from the propellant is the fundamental concept of the nuclear lightbulb.

Make it longer and bubble the reaction mass through the inert metal rather than just injecting at the center.

No, just duct the flow downward like the F-35.

I stand corrected. Didn't have a look at the high-res renders until now. That I'm not sure about. The possibility of using fewer tiles on the lower-temp areas is still decent. Maybe I was too hasty. It'as awesome to see and I'm glad. I just think that the renders show some lingering uncertainty, e.g., what the heck they're doing with the legs, and so forth. A thought occurred to me...if the Starship did start to list on landing (at least on the leeward side), there could be a computer subroutine to immediately flap the wings back in that direction to catch it. I would have to do some math to see whether it could rest in that position or not. Might be just enough if the PDSes were used sacrificially. Come to think of it, flapping them back could also help correct a windward list, by changing CoM.