37 thoughts on “Starship Landings”

  1. The strange thing for me is the number of self-styled experts still fighting Cluster’s Last Stand (hint to the history challenged: Cluster won). There are still folks who insist on the probabilistic nature of rocket engines, such that the calculated failure rate of multiple engines outweighs the calculated failure rate of a single engine, and therefore single-engine is safest. It’s usually the same people who insist if your rocket has engine-out capability, then it has too many engines.

    1. Why does the US Navy insist on two (2), yes count them — Two engines on their carrier aircraft? Something about it being a long swim if one of your jet engines fails flying over the ocean.

      2. Even before the F-35C, the Navy operated thousands of single engine aircraft. Even if we limit it to theVietnam era, there was the A-4, A-7, and F-8. There were countless more before that. The early Navy jet fighters had two engines because early jets had so little thrust. Most Navy fighters in the 1950s were single engine until the Phantom II.

      1. Dpn’t know what you’re asking, but I believe the RCS thruster blocks and the SPS were tied into the same fuel tanks, so in the event of SPS ignition failure, the RCS system coulf thrust long enough to reach TEI (which assumes it was post-landing and the LM engines were no longer available). I think the same goes for Orion.

        1. Thanks! I asked Rand specifically to comment because of his experience at North American Rockwell and if he had any insights he could share. Not that I had any contingencies in mind, but thanks for that. Yeah post landing with LM jettison for TEI is probably worst case for single point of failure. Interesting to note Orion has same capability.

  2. No Starship landing will ever be safe. Once they land, they’ll be in a country where crime is out of control, Antifa and BLM commit looting, arson, and murder on a whim, shoplifters can steal whatever they want, the overall murder rate is sky high, police are getting fired for not getting vaccinated, tens of thousands of unknown Afghanis are wandering loose, and violent drug and human trafficking cartels control the border.

    2. Once they land, they’ll be in a country where crime is out of control

      Yes very dangerous considering you be in Florida or Texas.

      But getting to the article he cites single stage starship for Point to Point I thought that was off the board. Didn’t think the starship with no booster is close to capable enough for worth while point to point like NY to Tokyo. The logistics of a NY to France/London or Cross continent just seem impractical.

      1. Yep. All the cartel members and gang members flooding across the Texas border are causing trouble, and lots of them are ending up in Florida, along with the Haitians.

        If you import the Third World, you become the Third World.

      2. Stupid politics aside,Musk has said Starship with no booster is capable of a 10,000km suborbital flight, with a different mix of engines. It’s always like this: who you going to believe, the spacecraft designer or some random boob on the Internet?

      3. “Didn’t think the starship with no booster is close to capable enough for worth while point to point like NY to Tokyo. ”

        Elon has stated Starship alone will be capable of 10,000 KM range on Earth, with a useful and salable payload. It just won’t do *anywhere* to *anywhere*, as long as you’re on Earth. The booster is needed for that.

    3. American civilians are still trapped in Afghanistan and American civilians are still leading the effort to get them out. Granted, there is some officially unofficial or unofficially official spook stuff going on too.

  3. There are other possible contingency profiles that I’ve excluded from this analysis, such as ones using all three RVacs w/ differential thrust and no SL engines.

    A pity. To me it would seem appropriate to use all three RVacs with de-rated thrust down to whatever is the appropriate setting and then gimbal with just one or two sea level Raptors (RSeas?). Then asymmetry in not encountered for heavy payload return so that compensation isn’t needed. Therefore one “RSeas” is plenty enough for control.

  4. I think they’ll have the engine redundancy well in hand. I think the larger remaining risk is the common failure modes in the tanks, tank pressurization, and plumbing. If there’s a critical failure there then no amount of good engines will help.

  5. Is there a scenario where one would have wished for a launch abort system that wouldn’t be solved by having solved all of the engine-out scenarios? With crewed rockers, for example, there’s the need to abort from the launch pad when the booster is exploding. Also, is there no risk of an exploding engine taking out other engines or rupturing the tank? Redundancy in case of engine out is nice but is that the totality of the risks?

    1. SN8: A pressure loss in the methane header tank resulted in a hard landing and loss of the vehicle.

      SN10: Helium ingestion caused a loss of thrust and a slightly hard landing, resulting in the loss of vehicle.

      SN11: A methane leak resulted in an engine explosion on restart which resulted in rocket parts raining down over a wide area.

      SN11 shows that there was an engine failure mode that blew up the whole thing, but there’s not much point in adding a lot of engine blast shielding to test vehicles, so that issue might go away.

      They’ve done fixes for the fuel flow and mixture problems, so let’s suppose those issues likewise won’t recur.

      I see that leaving some header-tank valve issues. If you have a common main valve from the header tank, it’s failure means you can’t ignite any engines and you lose the vehicle.

      If you instead relied on individual engine valves (after the main feed has branched off to individual engines), and no common header tank valve, then a valve that’s stuck open on one engine could drain all the fuel from the header tank, and you could lose the vehicle if the failure occurs long prior to relight.

      But that failure is unlikely because the valves should be closed during descent, and a valve isn’t going to open itself. The engine valves could also be paired in series so that it would take two failed valves to result in an uncontrolled leak, and then you could go to the series/parallel failsafe arrangement of four valves so that any one valve failure is okay.

      At that point your remaining physical risks (as opposed to software and controls) probably shift to re-entry issues, such as a loss of critical tiles, or failures in the flaperons, batteries, electrical systems, or tank pressurization systems (including COPV failures).

    2. All launch abort systems involve adding yet another rocket stage to the equation, thus increasing the risks, so you can always come up with some chance to total failure. The wings fall of your plane, you punch out, and if the ejection seat rocket doesn’t explode, your parachute fails…

      Musk’s plan is similar to the Shuttle: plan for it not to fail. The difference is, STS flew four times with tw-crew, then was declared operation. Musk us planning to fly Starship hundreds of times before declaring it opertional, and the first unknown number of launches will be without crew. It’s possible the first crewed Starships will be launched uncrewed, and then boarded in flight (from Dragon for dear.Moon, after orbital refueling; from Artemis or Gateway for lunar landing). P2P? Maybe when it’s more reliable than a 1950s jet airliner (and a lot better than the DeHavilland Comet…).

  6. The Blue Origin capsule landing with Bill Shatner in conjunction with this question makes me wonder if the crew and the rocket really must be paired all the time. Can’t the rocket land on its thrusters while the crew come down on the traditional tried-and-true capsule-under-a parachute?

    Sure, ride the rocket up. (Like THAT’s supposed to be “safe” …) but if we’re worried most about the ride down — don’t use that vehicle.

    1. But what if you’re not using SpaceX engineers and they forget to hook one of the parachutes to the capsule?

      If you don’t worry about pad aborts due to the booster’s massive redundancy, the Starship should be able to get itself into a glide mode throughout most of the envelope. That will get everybody down to a thousand or so feet in altitude and a pretty low velocity, in a nice stable orientation until the flip turn. For a small crew, that leaves some fairly good options for ejecting.

      I’ve mentioned before that landing horizontally (like a Harrier) wouldn’t be much of a challenge if you were willing to deal with the weight penalty of a different set of small engines.

    2. New Shepard’s capsule and rocket are not paired all the time. Post main engine cutoff, the capsule always separates from the rocket and descends separately. I think the capsule would make the mostly empty rocket too topheavy to land in one piece.

  7. A botched landing isn’t great. The debris field can be rather large. I assume they will work it out in the long run and it is a ways to go before they launch or land something crammed with people.

    A more important question, what happens if there is a mishap when they are testing in space refueling?

    1. It would depend on the mishap. The likelihood of an “explosion” blowing the Starship and tanker to a zillion pieces is low. Earliest plans call for the refueling to be done in the absense of crew. The whole point of the way Starship is being developed is to do everything unmanned until it’s shown to work reliably. People who claims this process will take years or decades are simply OldSpace apologists looking to continue their parade rain juju. I remember many discussions on NSF pre-2010, confidently predicting that any attempt to launch Falcon 9 would result in “a big explosion” because Musk would soon discover than running 9 engines simultaneously would not work. Remind me: has Tory Bruno withdrawn his statement that reusability woulld not wpork without “new physics?” Maybe he believe Musk is a space alien (rather than an immigrant from Africa alien)?

  8. For what value of ‘safe’?

    Are airline flights ‘safe’? How about other forms of public transportation – safe?

    Are automobiles safe? Is walking on a public street ‘safe’?

  9. The question is: How safe is safe? Commercial aviation is around 100 years old but the most dangerous years were probably right after WWII until the ’60’s. when air travel exploded using really crappy planes. As engine life went from tens of hours to tens of thousands of hours, fatality rates plummeted even though relatively few crashes are actually caused by engine failure.

    This only goes back to 1960 but the numbers are quite stark:
    https://www.bts.gov/content/us-air-carrier-safety-data

    I remember reading that the SME was supposed to be good fro a whole 300 or maybe 3,000 seconds and never came close. Being able to run an engine for hours indicates that the rest of the process is under control as well.

    I seem to recall that someone wrote a book about it.

  10. How much would it cost to buy a Starship? I heard that it would cost $212 million to build one. But I have read somewhere it would be $20 million, or less. Does anybody have the figures? I did read somewhere that it could be launched 1,000 times a year, and it could carry 850 passengers, and crew going point, to point.

    What about delivering 850 passengers, and crew to LEO? If so, then you could build orbital resorts up there. Say an orbital town with 3,000 residents, and 2,000 tourist.
    At 8,500 tons, I think Kalpana One would be great as an orbital resort.

    1. Likely the first one cost a quarter-mil, and the marginal cost now of building another, is $20 mil. The First 707 prototype nearly cost Boeing everything, the DC-8 ultimately cost Douglas Aircraft everything and the L1011 Lockheed Commercial.

      But, the last B707 was under $10-mil, iirc. And a far, far better product.

      1. I remember reading somewhere (don’t remember where) that the fabrication cost per Starship was $5mln, and that the initial group of engines ran $2mln apiece (so we’re talking $11mln per test article with 3 engines ea.). Musk has stated the target cost for mass produced Raptors is $250k apiece. Probably the main cost of a Mars-bound Starship will be the passenger-carrying guts. Does it matter how much they cost? So long as they come back in one piece, the cost gets amortized.

  11. Starship’s first orbit is going to crash/land in the ocean.
    And if launching and landing on ocean, that could be an abort option.

  12. There was a Falcon 9 that landed in the ocean and safed itself quickly enough that it didn’t explode when it tipped over. They wound up towing it back to Port Canaveral and lifting it out with a crane. Musk said maybe they’d launch it again if they could think of a reason, but they never did. I don’t know if it was scrapped or is somewhere waiting to become a museum piece. I think the biggest problem with safe abort during Starship launch is the 1200 tons of methane and lox on board. That’s a lot of wet stuff to dump in not much time. Then again, if the engines are all working, you could fly somewhere down range, I guess.

Comments are closed.