back to article Musk see: Watch SpaceX's latest Starship rocket explode while trying to touch down

The latest prototype of SpaceX's Starship rocket, the SN9, burst into flames as the vehicle attempted to land on Earth on Tuesday. All eyes were on the craft after its predecessor, the SN8, exploded during touchdown in December in Boca Chica, Texas. You can watch today's detonation in the video below. The accident occurs after …

  1. The Oncoming Scorn Silver badge
    Paris Hilton

    Flesh Gordon (Icon)

    Whenever I see these ships, this quote springs to mind from the above!

    Dr. Flexi Jerkoff: [Looking at his rocket ship] Fortunately I was able to complete the design before the Sex-Ray took effect.

  2. Sorry that handle is already taken. Silver badge

    I have to assume that SpaceX is pretty lucky not to have the book thrown at them for launching in violation of the licence after being explicitly told not to.

    1. This post has been deleted by its author

      1. diodesign (Written by Reg staff) Silver badge

        "so they stuck to the extant rules for the launch"

        That's not the way the FAA tells it. In an email to the media, the agency stated:

        "Prior to the Starship SN8 test launch in December 2020, SpaceX sought a waiver to exceed the maximum public risk allowed by federal safety regulations. After the FAA denied the request, SpaceX proceeded with the flight.

        "As a result of this non-compliance, the FAA required SpaceX to conduct an investigation of the incident.

        "All testing that could affect public safety at the Boca Chica, Texas, launch site was suspended until the investigation was completed and the FAA approved the company’s corrective actions to protect public safety. The corrective actions arising from the SN8 incident are incorporated into the SN9 launch license."

        This non-compliance is the reason why a January 28 test of the rocket was delayed, and why Musk went off on one on Twitter about it.


        1. Anonymous Coward
          Anonymous Coward

          Re: "so they stuck to the extant rules for the launch"

          Welp. I'm wrong again.

          1. awavey

            Re: "so they stuck to the extant rules for the launch"

            Not necessarily remember SN8 launched,very publically,December 9th. So we are saying the FAA started an investigation straight away but only notified SpaceX of it or this licensing problem that prevented a launch nearly 6 weeks later,as SpaceX didnt seem to be aware there were any problems,which was then all resolved within a matter of days,without SpaceX seemingly having changed,or only changed a little, the public protection protocols theyd been the fact they launched directly in violation it would appear of the FAA has then been quietly forgotten.

            Maybe the FAA have felt under pressure since their Boeing 737Max oversight to be seen as more a regulatory body flexing its power, than a mere rubber stamp.

        2. Grease Monkey Silver badge

          Re: "so they stuck to the extant rules for the launch"

          "As a result of this non-compliance, the FAA required SpaceX to conduct an investigation of the incident."

          Have the FAA learned nothing from the 737MAX "certification". They asked SpaceX to investigate themselves? Just like they asked Boeing to certify their own plane.

          Maybe this is a new model for all things. The Police could ask suspects to investigate themselves. Maybe we could all do our own MOTs and conduct our own driving tests?

          1. Anonymous Coward
            Anonymous Coward

            Difference here...

            SpaceX are developing a completely new class of vehicle and they expect it to crash a few times.

            They knew (from telemetry) what caused SN8 to fail before it even hit the ground, so there was no need for an investigation (as such).

            SN9 flew with a (temporary) fix to the problem that cause SN8 to fail. It crashed because one of the engines didn't start - I've not seen anything yet to say that the cause of that is known (and Elon is "Off Twitter for a bit", so the best source is currently not available).

            It may be that SpaceX need to do some investigating to understand what happened. I can hardly see the FAA getting involved in this (at least, not in depth) as they are very unlikely to have anyone able to comment on what is a novel platform.

    2. WonkoTheSane

      They DID get permission to launch from the FAA the day before.

  3. Winkypop Silver badge

    Never mind the landing

    How about that take off!

  4. Gene Cash Silver badge

    Did not explode in the air

    "unfortunately, it exploded in the air before it could flip the right way up and touch down"

    Did you not watch the video? It explodes on contact with the ground.

    1. Lorribot

      Re: Did not explode in the air

      Definitely hit the ground, also the commentary said they would ignite two engines to break and rotate, looks to me like the second one failed to fully ignite, there was brief burst but not clean then just the one engine so I assume it didn't slow sufficiently then it was still at an angle.

      They may need to rethink that relighting sequence as there is not enough time to ignite another engine should one fail, would be better to light three engines initially then kill one to reduce power and maybe get it upright earlier, all seems to be cutting it a tad fine.

      1. Notas Badoff

        Re: Did not explode in the air

        Turn it off, and turn it on again. But earlier next time, okay?

        1. Anonymous Coward
          Anonymous Coward

          Re: Turn it off, and turn it on again

          Everyone's an expert when it comes to rocket science

          1. Anonymous Coward
            Anonymous Coward

            Re: Turn it off, and turn it on again

            Thing is, El Reg is one of those few places where you will actually find a rocket scientist!

      2. John Robson Silver badge

        Re: Did not explode in the air

        The difficulty is that if you do that then you reach zero velocity whilst still having positive altitude.

        The F9 1st stage is completely unable to hover, I imagine that starship will be able to hover dependant on cargo, but it's a serious waste of propellant to do so.

        When was the last time they failed to land a booster (i.e one that they tried to land within their normal envelope)?

        How many attempts did it take to stick that first one?

        This is a very early test vehicle, it has had substantially more real world testing than SLS, they're just trying to do things that are actually hard.

        The one remaining engine looked as though it had good fuel pressure all the way, mach diamonds almost all the time - so I guess it wasn't a fuel pressure issue this time, but something else in the raptor lighting sequence.

        1. Andy The Hat Silver badge

          Re: Did not explode in the air

          Starship in it's current configuration is also unable to hover. As I understand it the first engine will flip and slow the vehicle, the second is used almost as a landing retro as two engines cannot be throttled back enough to sustain hover. So it's rock and a hard place - plummet vertically with little drag or directional control and use loads of fuel trying to control both the speed and position or, go for the high drag, high(er) aerodynamic control and little fuel use then do the more complicated flip and stop manoeuvre.

          Sounds easier to do the first (and they do it enough on orbital missions) but when fuel load equates to payload and payload is the premium asset, the second option could have massive payload and thus economic advantages.

          1. John Robson Silver badge

            Re: Did not explode in the air

            I think (given the under rotation of this test) that they want both engines for the flip manoeuver and for the landing. They certainly manage to hover at 10km, This isn't expected to take off from low altitude, it only has to land at low altitude - when it is firing for propulsion it will be doing so with the fixed vacuum engines (not yet installed) since it will already be pretty high up.

            (Mars low atmosphere is pretty low pressure, so the vac optimised engines will be used there as well).

            The equations change again when you still have cargo on board (I would expect a hover to be just about technically possible, even if deeply not desirable).

            The F9 does as much drag as possible as well, but it isn't designed for anything other than first stage recovery (as per starship booster, which I think also returns vertically). The starship (upper stage) is going to be returning from orbital (or hyperbolic) speeds, and that is orders of magnitude more energy to deal with. I do wonder if they might try to do the flip slightly earlier, but then they'd need grid fins as well as the air brakes/flaps/elonerons that they currently have - and it would be less efficient overall as well.

            Rocket science is easy, it's the engineering that's hard.

            They have a completely different approach to virtually any other rocket company - they test in the real world rather than just doing calculations.

            They fail, and learn, fail and learn...

            They've gone from this to (I think) five landing failures out of the 80+ launches since 2017

    2. Gordon 10

      Re: Did not explode in the air

      Its OK. Musk has ordered 10 vintage Lunar Lander arcade games for the SpaceX canteen.

    3. macjules

      Re: Did not explode in the air

      To Boeing that would be a 99% success.

  5. aregross


    It appears only one engine successfully refired, and even that one didn't seem up on cam.

  6. John Jennings

    Looks like the second engine didnt start properly, though I am no rocket scientist... Pity - appeared to be good up to the end. Of course its the end that matters!

  7. JassMan

    Maybe it needs a clunk tank

    Model aeroplanes and I suspect some proper aerobatic ones are fitted with a small tank containing a weighted (clunk) flexible pipe which alwsys gets fuel from the deepest bit on the tank regardless of orientation. Maybe the SNxes suffer from the fuel taking too long to settle at the notional bottom of the tank and the motor is trying to burn gas rather than liquid fuel. Presumably it would need the same arrangement for the oxidant.

    I don't suppose Elon will ever read thisthough.

    1. eldel

      Re: Maybe it needs a clunk tank

      I'm fairly sure I read that (one of the points of) SN8 was to test the pressurised header tank. This one talked about the landing tanks. I would have assumed that they would also be pressurised given that they have to supply fuel from a 90 degree (ish) angle to allow it to align for landing.

      Whatever the reason I suspect that we'll find out about it in a few days. Muskies girls and boys seem to be very open about what went wrong. Of course - when you aren't having to justify another $XB from congress that sort of thing is probably easier.

      1. dankell

        Re: Maybe it needs a clunk tank

        Last time the failure was due to lack of pressure from the head tank.

        This time they pressurised the head tank with hydrogen but did state that that was a temporary solution.

        You can definately see the second rocket flare and try and light.

        1. Anonymous Coward
          Anonymous Coward

          Re: Maybe it needs a clunk tank

          I think it was helium ;-)

  8. Anonymous Coward
    Anonymous Coward


    "Landed" a bit close to SN10. Hope there was no damage.

    1. John Robson Silver badge

      Re: SN10

      Video really foreshortens that distance.

      It wasn't nearly as close as it looked (which is good)

    2. genghis_uk

      Re: SN10

      I was thinking it is a bit brave to have both rockets on their launch pads at the same time.

      Even with a large gap, if the tracking went slightly astray, they could have an even more expensive clean up operation - and very little scientific data from SN10

      1. Eclectic Man Silver badge

        Re: SN10

        Yes, you have to admit though that it was a very accurately navigated crash.

        1. Stuart Halliday

          Re: SN10

          A good crash is one where you can get all the pieces? :)

          1. Anonymous Coward
            Anonymous Coward

            Re: SN10

            And a good landing is one you can walk away from.

  9. Spherical Cow

    Asleep on the job

    The countdown person missed the 10 second mark and then had to rush the countdown to catch up.

    1. Matt_payne666

      Re: Asleep on the job

      you had one job... one f**king job... do the countdown... every one knows to launch a spaceship you count down from 10... and you couldnt even get that right!!!

  10. yetanotheraoc Silver badge

    Billionaire rocketman

    When I first heard about the pay for seats plan, I figured it would be something like three professionals and one tourist. Recent general media reports make it sound like four tourists.

    "Issacman, who is qualified to fly military jets ..."

    Thanks for that detail, ElReg, missing from other reports. It makes me feel a *little* better about the planned mission. Still, what about the other three warm bodies, do they have any qualifications? Or will the rocket be relying mostly on the perfected self-driving technology?

    1. Anonymous Coward
      Anonymous Coward

      Re: Billionaire rocketman

      From the application form

      94 Do you have any desire to die in a massive fireball? Y/N

      Y - proceed to next question, N - thank you for your enquiry, but unfortunately

    2. Dave 126 Silver badge

      Re: Billionaire rocketman

      Iscaacman will be the commander of the vessel, which in this case mostly means he has authority over the other occupants. The craft itself is controlled from the ground or by its computers.

      The only real control the commander has is the big red lever which will, in the event of Something Bad happening during the launch, blast the Dragon capsule safely away from the launcher.

      1. Jellied Eel Silver badge

        Re: Billionaire rocketman

        The only real control the commander has is the big red lever which will, in the event of Something Bad happening during the launch, blast the Dragon capsule safely away from the launcher.

        I suspect even that is just an illusion of control. The computers can almost certainly detect an impending UED faster than the capsule commander. It's also a test of any Dragon* AI, and if that AI decides it's safer to eject the warp core.. I mean crew, than launch.

        *Any fan of Neal Asher knows that Dragon works in mysterious ways.

        1. LogicGate Silver badge

          Re: Billionaire rocketman

          Knowing a bit of what is required for flight-software, I am quite convinced that there is no AI or neural network involved in the dragon / falcon software decision making. Happy to be corrected though.

      2. sabroni Silver badge

        Re: blast the Dragon capsule safely away from the launcher.

        And directly at the nearest concrete off ramp divider?

  11. mevets

    Tesla: Data Survivalists.

    I think that is a good corporate logo; and maybe replace the anonymous-eyebrow-nose logo with an exploding spaceship.

    1. Spherical Cow

      Re: Tesla: Data Survivalists.

      It's not an eyebrow-nose, it's part of an electric motor:

  12. a_yank_lurker

    Not Unexpected

    Problems with the rockets are not unexpected especially when doing something relatively new. The key is identifying and fixing the problem which his Muskiness seems to be willing to do.

  13. Neil Barnes Silver badge

    The bad news: oops!

    The good news - there's probably enough left to find out what went wrong. "We've just got to work on that landing a little bit..."

    Is it just me, or do the youtube videos look more like a 50's sci-fi film set than real life?

    I think the engineers need one of these -->

  14. Totally not a Cylon

    Engine problem

    Looks to me like the vehicle is working as intended.

    Just the Raptor engines have a problem, which is not surprising as they are a new design of 'Full-Flow Engine'.

    So the question now is: Do they continue to basicly discard a launch vehicle each time they test or do they built a test stand for 3 Raptors and perform mission profile start/stops?

    (Remember that Mercury/Gemini/Apollo went through the same 'Vehicle explodes during testing' as did Falcon but that now 'sticks the landing'.....)

    Which option gives SpaceX the data they need?

    Whilst also fitting what Elon wants to do? (The very public testing clearly separates SpaceX from the others)

  15. trevorde Silver badge


    "rapid unscheduled disassembly" = explosion

  16. First Light

    Twitter break

    You bet he's off Twitter. His lawyers are getting ready for the SEC phone call.

  17. Dave 126 Silver badge

    "We don't like the term 'meltdown'. We prefer 'unrequested fission surplus'. " - Mr Burns

    1. Yet Another Anonymous coward Silver badge

      "Engine rich burn" in the previous test.

  18. Schultz

    "Work a little bit on the landing"

    I guess the commenter of the video nailed it. Don't board that ship before they did their little bit of work!.

    1. Yet Another Anonymous coward Silver badge

      Re: "Work a little bit on the landing"

      I don't believe the plan is to land the crew compartment on the booster anyway - so this would still have been a perfect launch, even with the 3 victims/idiots/customers aboard

      1. Anonymous Coward
        Anonymous Coward

        Re: "Work a little bit on the landing"

        Starship is the "crew compartment" - it will be sitting on top of an even bigger booster when it's launched to orbit - shows the various configurations and the Super Heavy booster they will sit on.

  19. Steve Kerr

    I wonder...

    Did it ask the ground whether it would be friends?

    Kudos for actually running and testing prototypes to failure, can learn much from failure as from success, all learning experiences with lots of data during failure.

  20. katrinab Silver badge

    Put a human in it later this year?

    I have no doubt that the data they obtained from this test will help them improve things, and they will sort out the problem(s) eventually. But even if the next launch is completely successful, I would want to see a lot more than 10 months of successful launches and landings before going in it myself.

    1. Anonymous Coward
      Anonymous Coward

      Re: Put a human in it later this year?

      There's no way they will be ready this year - lots of development to do yet.

      1. Anonymous Coward
        Anonymous Coward

        Re: Put a human in it later this year?

        Someone thinks they will? Even if they were ready now, it takes a long time (for good reasons) to get human flight rating certification.

  21. heyrick Silver badge

    As I said to my oncologist when my mother died...

    The failures are every bit as important as the successes. You need to know not only what does work, but what does not. And, hopefully, why.

    SpaceX might have a reputation for crowd pleasing explosions, but every single one will be providing masses of data and telemetry, good examples of what didn't work in order to make better things in the future.

    Icon, because I watched the video...

  22. TeeCee Gold badge
    Black Helicopters

    I reckon they did it deliberately.

    This way they can now go to the FAA, thumb their noses and say; "Look. Doing it by your rules makes no sodding difference, so exactly what purpose do you bunch of overpaid, clipboard wielding tossers serve?".

    It's what I'd have done.

    1. Jellied Eel Silver badge

      During the fuel loading phase, there was an unexpected pickup in the danger area. So the FAA may want to know how that happened. There didn't appear to be any pursuit, but if the FAA demands more security to prevent that happening again, I guess it'll get more complex/expensive.

  23. Grease Monkey Silver badge

    Didn't look like it exploded when trying to land to me. Looked more like it crashed into the ground and exploded on impact. Not a particularly fine distinction.

  24. mevets

    Enhanced Widow Maker

    This was actually a test of a new software add on for the model S. It is a bit pricey, but with the insurance payout, your SO will be able to afford it.

  25. Trilkhai

    I wonder how much of an impact all of this has on global warming?

    Thinking about that as I noticed this headline right after hearing more talk about spending tax money to give hefty earned-income tax credits to people who can justify buying a brand-new electric car, thinking about the extra taxes/fees added in some countries/states to purchases of gas/petrol, electronics, cans, etc., non-carpool drivers being stuck in heavier traffic so the electric-car drivers can get their own special lane, and so forth. It's a bit annoying to consider that we might be making all of this effort only to have some (all?) of the impact wiped out by things like this...

    (Don't even go down the "we need to colonize other planets so we have somewhere to go after trashing the Earth" road, considering that's on par with "I need to buy a janky roach-infested apartment so I have somewhere to live when my habit of setting the furniture on fire causes my $900,000 house to burn down.")

  26. Anonymous South African Coward

    Optional seats

    ....said he will be donating the other three seats to chosen members of the general public

    Can we nominate who to put on those seats?

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like