9 ms·

Illustration how "move fast and break things" in aerospace domain works.

ericd 6y ago
Honestly, my main takeaway from this testing and Falcon 9 dev has been that we should be doing a whole lot more of that than we have been, at least until there are human passengers. The contrast in rate of progress with eg SLS is stunning.
- api 6y ago
  Also keep in mind that it's already doing what pre-F9 rockets did: go up. The RUDs so far are on landing attempts. Landing is a whole lot harder.
  - unanswered 6y ago
    Unfortunately it seems like traditional aerospace now has a very big lever to pull to get SpaceX shut down: public opinion following the reasoning "rocket go boom. go boom bad. rocket bad".
    - m4rtink 6y ago
      Or actually the other way around - what will be in the news, another successful Atlas V or Ariane 5 or a nice big powerful explosion during a test flight ? That can bring a lot of interest from people not to mention disseminate the news about Starship far and wide.
    - api 6y ago
      If that was going to work it would have already. SpaceX has been blowing up rockets for years.
  - simonh 6y ago
    There's a key difference though, Starship is intended to carry people. The correct comparison for this isn't landing Falcon 9 boosters, it's landing crew capsules.
    - Cogito 6y ago
      Falcon 9 was also intended to carry people. Starship is a more capable upper stage then Crew Capsule, part of which means that it can launch itself (at least suborbital). They aren't trying to qualify these prototypes as human-rated, they have to finish the design work first! For this pase of testing it's much more accurate to compare to Falcon 9 than Crew Capsule.
      - simonh 6y ago
        Oh come on, the Dragon capsule brings people back to the ground, not the first stage booster and it's landing these things that is the issue so far, not launching them. All I am saying is that this vehicle, in it's final qualified form, will be landing human beings on the deck back here on Earth. That's the objective, so it's viability and reliability needs to be considered in that context. For an F9 first stage a failure rate of 1/10 is fine. One out of 20 is outstanding. For Starship, these things are intended to carry a hundred paying passengers or more. The failure rate needs to be lower than 1 in a thousand. Much lower.
        Cogito 6y ago
        I mostly agree with you, but I'm not sure on why you're focusing human passenger safety on this at this early development stage, as if it's somehow a reasonable thing to focus on. Launching these things IS the issue so far, by far the main objective of these tests. Landing them would be nice, but the primary objectives are to launch the whole thing (in particular including aero surfaces and multiple engines, as compared to previous prototypes) and test the various novel maneauvres that they are attempting. That almost all the objectives are being met in these tests is a testament to how good SpaceX is getting at launching flying water tanks, and how hard the entire flight profile is to achieve (launch, belly flop, landing flip, landing). In the anticipated human-rated version of Starship safety will be immensely important, but this is nowhere near being that thing! This was only the second prototype that even started to look like 'Starship'. We have no idea what a human-rated version would loook like, and a human-rated Starship is definitely not the objective of these tests. The Starship prototypes they are testing at the moment are much much much closer to a Falcon 9 than a capsule. Specifically the systems and processes they are testing all have fairly direct parallels in Falcon 9, and almost no correspondence to a capsule. For example, Falcon 9 and Starship both have: super chilled propellent tanks, multiple large rocket engines, active flight control systems, engine relight, active aero surfaces, and powered landing. Starship and the capsule both have... heatshields? Human passenger safety will be immensely important for Starship, but to focus on it at this stage of development would be immensely premature.
- codefreakxff 6y ago
  Well. Yes. It appears to be a fully functioning launch vehicle and just as reusable as its predecessors, don’t see much downside in making the landings more “interesting”. When they work out the dynamics of it then they really have something that will revolutionize launches
- Already__Taken 6y ago
  We knew this since watching how much the Russians got done in the time they had for the resources. Put the engineers together with the manufactures, build often, test more.
  - wp381640 6y ago
    just don't sit near the launchpad
    - krisoft 6y ago
      And if anyone believes wp381640 is joking, this lesson seems to have been learned the hard way: https://en.m.wikipedia.org/wiki/Nedelin_catastrophe https://en.m.wikipedia.org/wiki/Nedelin_catastrophe
      - Breza 6y ago
        "54–300 deaths (exact number not known)" Wow, I can't imagine a disaster happening in the United States where we barely even knew the order of magnitude of fatalities.
- kiba 6y ago
  The way to make them more reliable is not to put human passengers while increasing the iteration rate and manufacturing speed. The two ways I heard to get really reliable products is to get good at manufacturing them reliably at scale or be really slow and really careful in making them one at a time.
MR4D 6y ago
‘We may crash, but at least we do it with style!” Somehow that should be an Elon tweet.
- jerf 6y ago
  If you happen not to have seen it, "How Not To Land an Orbital Rocket Booster": https://www.youtube.com/watch?v=bvim4rsNHkQ https://www.youtube.com/watch?v=bvim4rsNHkQ from SpaceX.
walrus01 6y ago
They landed a lot of falcon 9 first stages without legs on the water, vertically, to test the control system. And a number of attempted barge landings with legs failed, before the first successful landing. Now it's almost a routine thing.
- kevin_thibedeau 6y ago
  That's probably what FAA was peeved about SN8. They might have been expecting a water landing and didn't like that an unproven rocket was sent back to dry land.
  - simonh 6y ago
    No, it was the last minute engine swap.
    - twicetwice 6y ago
      Is there a source for that? I've heard tons of speculation & I'd be very interested to know if the mystery has definitively been settled!
  - Laremere 6y ago
    The Falcon landed on land before it landed on a barge. Obviously landing on a moving/swaying barge is harder, however they had to start with water landings because they hadn't proved they could aim at their target. Once they crashed into the barge a few times, the FAA was convinced they'd only blow up the landing pad and not drop it on something else in the launch complex. So they got clearance for landing on land. The early starhopper/starship flights proved what they needed to that they'd generally land on target.
lrossi 6y ago
You could also call this TDD.

Honestly, my main takeaway from this testing and Falcon 9 dev has been that we should be doing a whole lot more of that than we have been, at least until there are human passengers. The contrast in rate of progress with eg SLS is stunning.

api 6y ago
Also keep in mind that it's already doing what pre-F9 rockets did: go up. The RUDs so far are on landing attempts. Landing is a whole lot harder.
- unanswered 6y ago
  Unfortunately it seems like traditional aerospace now has a very big lever to pull to get SpaceX shut down: public opinion following the reasoning "rocket go boom. go boom bad. rocket bad".
  - m4rtink 6y ago
    Or actually the other way around - what will be in the news, another successful Atlas V or Ariane 5 or a nice big powerful explosion during a test flight ? That can bring a lot of interest from people not to mention disseminate the news about Starship far and wide.
  - api 6y ago
    If that was going to work it would have already. SpaceX has been blowing up rockets for years.
- simonh 6y ago
  There's a key difference though, Starship is intended to carry people. The correct comparison for this isn't landing Falcon 9 boosters, it's landing crew capsules.
  - Cogito 6y ago
    Falcon 9 was also intended to carry people. Starship is a more capable upper stage then Crew Capsule, part of which means that it can launch itself (at least suborbital). They aren't trying to qualify these prototypes as human-rated, they have to finish the design work first! For this pase of testing it's much more accurate to compare to Falcon 9 than Crew Capsule.
    - simonh 6y ago
      Oh come on, the Dragon capsule brings people back to the ground, not the first stage booster and it's landing these things that is the issue so far, not launching them. All I am saying is that this vehicle, in it's final qualified form, will be landing human beings on the deck back here on Earth. That's the objective, so it's viability and reliability needs to be considered in that context. For an F9 first stage a failure rate of 1/10 is fine. One out of 20 is outstanding. For Starship, these things are intended to carry a hundred paying passengers or more. The failure rate needs to be lower than 1 in a thousand. Much lower.
      - Cogito 6y ago
        I mostly agree with you, but I'm not sure on why you're focusing human passenger safety on this at this early development stage, as if it's somehow a reasonable thing to focus on. Launching these things IS the issue so far, by far the main objective of these tests. Landing them would be nice, but the primary objectives are to launch the whole thing (in particular including aero surfaces and multiple engines, as compared to previous prototypes) and test the various novel maneauvres that they are attempting. That almost all the objectives are being met in these tests is a testament to how good SpaceX is getting at launching flying water tanks, and how hard the entire flight profile is to achieve (launch, belly flop, landing flip, landing). In the anticipated human-rated version of Starship safety will be immensely important, but this is nowhere near being that thing! This was only the second prototype that even started to look like 'Starship'. We have no idea what a human-rated version would loook like, and a human-rated Starship is definitely not the objective of these tests. The Starship prototypes they are testing at the moment are much much much closer to a Falcon 9 than a capsule. Specifically the systems and processes they are testing all have fairly direct parallels in Falcon 9, and almost no correspondence to a capsule. For example, Falcon 9 and Starship both have: super chilled propellent tanks, multiple large rocket engines, active flight control systems, engine relight, active aero surfaces, and powered landing. Starship and the capsule both have... heatshields? Human passenger safety will be immensely important for Starship, but to focus on it at this stage of development would be immensely premature.
codefreakxff 6y ago
Well. Yes. It appears to be a fully functioning launch vehicle and just as reusable as its predecessors, don’t see much downside in making the landings more “interesting”. When they work out the dynamics of it then they really have something that will revolutionize launches
Already__Taken 6y ago
We knew this since watching how much the Russians got done in the time they had for the resources. Put the engineers together with the manufactures, build often, test more.
- wp381640 6y ago
  just don't sit near the launchpad
  - krisoft 6y ago
    And if anyone believes wp381640 is joking, this lesson seems to have been learned the hard way: https://en.m.wikipedia.org/wiki/Nedelin_catastrophe https://en.m.wikipedia.org/wiki/Nedelin_catastrophe
    - Breza 6y ago
      "54–300 deaths (exact number not known)" Wow, I can't imagine a disaster happening in the United States where we barely even knew the order of magnitude of fatalities.
kiba 6y ago
The way to make them more reliable is not to put human passengers while increasing the iteration rate and manufacturing speed. The two ways I heard to get really reliable products is to get good at manufacturing them reliably at scale or be really slow and really careful in making them one at a time.

MR4D 6y ago

‘We may crash, but at least we do it with style!” Somehow that should be an Elon tweet.

jerf 6y ago
If you happen not to have seen it, "How Not To Land an Orbital Rocket Booster": https://www.youtube.com/watch?v=bvim4rsNHkQ https://www.youtube.com/watch?v=bvim4rsNHkQ from SpaceX.

walrus01 6y ago

They landed a lot of falcon 9 first stages without legs on the water, vertically, to test the control system. And a number of attempted barge landings with legs failed, before the first successful landing. Now it's almost a routine thing.

kevin_thibedeau 6y ago
That's probably what FAA was peeved about SN8. They might have been expecting a water landing and didn't like that an unproven rocket was sent back to dry land.
- simonh 6y ago
  No, it was the last minute engine swap.
  - twicetwice 6y ago
    Is there a source for that? I've heard tons of speculation & I'd be very interested to know if the mystery has definitively been settled!
- Laremere 6y ago
  The Falcon landed on land before it landed on a barge. Obviously landing on a moving/swaying barge is harder, however they had to start with water landings because they hadn't proved they could aim at their target. Once they crashed into the barge a few times, the FAA was convinced they'd only blow up the landing pad and not drop it on something else in the launch complex. So they got clearance for landing on land. The early starhopper/starship flights proved what they needed to that they'd generally land on target.

lrossi 6y ago

You could also call this TDD.