Starship Abort Modes

What will Starship do if bad things happen during flight?
This video explores abort options on four different crewed vehicles to better understand how abort might work for Starship
@Eric_Gunnerson on Twitter
Triabolical_ on Reddit

[@jgottula]

I thought it was a nice touch to use the hospital “scale of pain” chart faces for the various abort scenarios. Love it! 😄

[@EagerSpace]

I'm glad that somebody picked up on that...

[@NikanDragosysSerpenDra]

@@EagerSpace NOTICED TOO

[@Kieran2350]

"If you liked this video bake yourself a cake" is the best ending to a yt I've heard. Looking forward to the one on parachutes. There is some youtube video out there about how Origami got popular after they realized every shape could be folded from a single sheet of paper, maybe that's related.

[@dsdy1205]

11:33 Depending on how early in the launch the Starship aborts, it may never even need to perform a belly flip. If that's the case, then it only needs to burn off enough fuel to ensure that it can land with however many engines out the design team desires, and hover straight to a landing.

[@EagerSpace]

Really interesting point and I'm unhappy that I didn't think of it.

[@dsdy1205]

@@EagerSpace well, it's not watertight; given the bellyflop requires some distance to fall to establish stable attitude control, there may be a point where you're too high up to do a full hover landing, but too low to do a belly flop. I haven't done the math to see if this regime exists, but in that case the crew is doomed.

[@alkimball8920]

If a heat tile crippled starship were able to reach LEO or something close, perhaps a rescue mission from a waiting dragon capsule could be quickly launched. I realize this would require additional docking technology for starship and if I were Space-X I would always have a falcon-9 or two or three in waiting for this very purpose (depends on how many people you're launching on Starship). Perhaps the emergency docking mechanisms could be optomized for weight and mass on the starship side of things to keep added weight down. I think it's worthy of consideration when you remember the loss of NASA's 2nd shuttle STS-107... If they knew what they were facing on re-entry they could of and should of aborted. Same goes for Starship I believe. I would call this a last ditch "mission abort" but when you consider the deadly alternative, it's worth doing.

[@EagerSpace]

The simplest thing would be to launch another undamaged crewed starship and pick the people up.

[@J7Handle]

@@EagerSpace While we're talking about the possibility of Crew Dragon being used to rendezvous with Starship, is it possible Starship HLS could be certified to carry astronauts in space before any Starship certifications to launch or reenter with humans onboard? Because I can envision scenarios where, as a stopgap measure, HLS ships are launched to LEO unmanned, then acquire crews via rendezvous with Crew Dragon, and then the return mission having another rendezvous to switch the crew back to Crew Dragon for reentry.

[@engineerstudentshenanigans3452]

Very thorough, very informative video! I am concerned about the Vacuum Raptors being used in an abort since SpaceX was talking about how they'd be very Overexpanded at Sea Level. Hopefully not destructively so.

[@EagerSpace]

They test the vacuum raptors at sea level at McGregor, so they are good enough for that usage. kzbin.info/www/bejne/pYW0qqCZiJ10pqc You can see effect of the higher pressure pinching the plume right at the edge of the nozzle, and the formation of a mach diamond very close to the nozzle.

[@engineerstudentshenanigans3452]

@@EagerSpace I was under the impression SpaceX started that those tests were using a shorter, lower expansion nozzle than what was intended for Production VacRaptors. I could be mistaken though.

[@807800]

Haven't watched it yet, but let me just say, thank you for the video!

[@ksmi9109]

My question is for Starship’s abort capabilities late into Superheavy’s burn assuming a majority of it’s engines are still operational and the booster is intact. Would a fully fueled Starship with payload be able to accelerate faster than an unburdened Superheavy booster, or at least at a significant enough rate to get out of it’s way? That puts a pretty wide black zone in Starship’s abort capabilities.

[@EagerSpace]

I think the answer to your question is "no". 33 raptors will give super heavy plus starship more acceleration than 9 raptors on a fully loaded starship. But the scenario seems unlikely to me. Why would you want to abort from Super Heavy if all the engines were still running, and if you did, what would keep starship from being able to shut down the engines on super heavy?

[@ksmi9109]

@@EagerSpace yeah, I figured as much a bit later as I was thinking about it. It isn’t a very likely scenario. Thanks for the response 👍

[@OzearEimaj]

Assuming all the Booster engines continue to fire after separation, this is not survivable if the Ship is fully fuelled. However, if you under-fuel the Ship (as Musk has suggested to make pad aborts possible), then this is possible: 50% filled Ship: can abort up to 35s into flight if booster continues to fire (LEO capacity reduced to 50t) 30% filled Ship: can abort up to 90s into flight if booster continues to fire (LEO capacity reduced to 20t)

[@topsecret1837]

@@OzearEimaj Also not to mention that since 80% of the propellant mass is LOX the ship can dump almost all of that very quickly.

[@topsecret1837]

@@OzearEimaj This also assumes that the booster cannot have a sudden increase in drag, for which it can for very obvious reasons in order to slow down on nominal entry and to drop below sonic speed for landing. They would need to be angled though to allow drag to be produced in ascent whereas a 0° AoA would not produce drag in that phase of flight.

[@macieksoft]

Don't worry, before we get crewed Starship we gonna have so many Starlink sats so the biggest concern will be colliding with a satellite and not the failure during boost ;-)

[@sidharthcs2110]

Starship looks like a death trap

[@algodoomarbleracing]

1:46 No pain 2:01 Mild 2:40 Moderate 5:23 Severe 6:35 VERY SEVERE 6:58 *W O R S T P A I N E V E R*

[@EagerSpace]

I'm so happy some people picked up on that.

[@colten1825]

@@EagerSpace i was too focused on the dog in the pool

[@LeonelEBD]

Sometimes I think ... If you're going to use SRBs for launching people, then go big and use an AJ260 already. Anyways if the smaller ones expldoes everyone's is dead anyways. So let's just go cheaper. Thank God I'm not in congress =D .

[@Lord_P-NT]

Sound like Ares I

[@kittyyuki1537]

2023 Update, Starship will be using Hot Fire Staging for stage separation, with an addition of a vented interstage on top of Super Heavy with thick steel bulkhead to protect the top. That makes Starship abort more feasible. There is a new Raptor in the works, Raptor 3 currently in testing managed 270 tons of thrust so far.

[@paulloveless9180]

Nice. Exactly the type of content I love. I hope your channel gets more exposure.

[@Makoto778]

Good video, but few more things to note... 8:13 Shuttle kinda* had a pad abort but it involves flinging the hatch open, jumping into the Emergency Egress Basket (basically a zip line) and getting to a bunker 1/4 mile away, all while hoping nothing explodes. The same baskets were kept for Dragon and Orion, and are to be used for less urgent emergencies. 9:27 The danger of the SRBs probably partially explains the high g's for the Orion abort. Also worth noting the the SRBs on Shuttle/Orion used multiple steel segments, while most SRBs on Delta/Atlas/etc used composite/fiberglass materials. Steel usually doesn't fail dramatically, but gradually weakens, deforms and leaves gaps for hot gas (as in the case of Challenger). Contrary, single piece fiberglass/composites can crack easily, and a small crack will quickly expand and cause the whole booster to explode. Also, Just for fun, the Ares I/Orion would have had the second worst abort capability (with shuttle being the worst) because it would parachute though the cloud of burning SRB fuel during any abort within 30-60 sec of launch. Lets hope not such issues are present for SLS/Orion. (images.spaceref.com/news/2009/fratricide.report.pdf)

[@EagerSpace]

Thanks. One note: news.northropgrumman.com/news/releases/northrop-grumman-and-nasa-successfully-test-space-launch-system-booster

[@J7Handle]

"Landing anywhere" with a crewed Starship is going to require landing legs to land on land, and probably something to cushion the tip over when landing in water. My estimates are that the top of the ship hits the water with around 60 mph of velocity when tipping over. My best guess at a solution is a few small SRBs to arrest the tipping motion right before it's fully horizontal. That would require the ship to fall on its "back", or non-heatshielded side, which is fine assuming the crew don't end up hanging out of their seats upside down. But it's easy to envision a seating arrangement that suits three orientations of normal acceleration. Just need seats that can rotate 90 degrees. Overall, a "land anywhere" abort mode for Starship would probably take around 5 tons of landing legs and SRBs to accomodate, as well as a suitable seating arrangement and means of disembarking when the ship is on its back in the water.

[@seagie382]

10:35 Hot staging, baby!

[@edki669]

Orion and Dragon are the only capsules capable of aborting after s structural failure, like the one seen on that crewed Soyuz launch not that long ago

[@EagerSpace]

Why do you think Starship is unable to abort in that scenario? BTW, the Soyuz failure on MS-10 was a staging failure that led to structural failure in the second stage.

[@edki669]

​@@EagerSpace In case of a staging failure, there are no aborts available for Starship or the Shuttle, since they are essentially the second stage. Granted, seperating Orion from the ESM is also a staging event, however, it takes place in space after orbital insertion, or TMI, in this case, which means you have much more time at hand to troubleshoot the issue. During launch, you're toast.

[@EagerSpace]

@@edki669 Ah. You said "structural failure" in your original response... Staging failures in general do tend to be quite problematic - that's why SpaceX has gone with pneumatic systems that they can test - but there are certainly cases where if you can't get starship disconnected from Super Heavy, you will have a very bad day. There are also cases where if your abort system on Orion malfunctions - if you cannot jettison it - you simply cannot reenter successfully. All the systems out there have tradeoffs.

[@topsecret1837]

@@EagerSpace The booster already has four very high drag producing surfaces which if angled could also provide drag while ascending. A combination of that with trying to push the ship of the booster via gas pressure like some missile silos did to throw the door off to the side would solve the problem of separation.

[@EagerSpace]

@@topsecret1837 If Starship/Super Heavy stage at about where Falcon 9 does - around 50km or higher - there's pretty much zero air left and therefore no real drag. Somewhere down lower there's an altitude where that's no longer true, and then it's the drag of the grid fins versus the drag of the nose cone. The nose code is going to have more drag based on area but the grid fins are more draggy. Not sure how that one comes out.

[@w0ttheh3ll]

If starship emergency lands in the ocean, won't it fall apart upon toppling over? Even if not, the crew won't like it much ...

[@EagerSpace]

It's fairly robust and light so it should be fine in light seas, but it could break up in heavy seas. Crew definitely will be unhappy.

[@kekkeyan8614]

I feel like starship pad abort will be like shuttle rtls. requiring continuous miracles interspersed with acts of god to work. Could 9 raptors start up fast enough and have enough thrust at SL to push a fully fueled starship away from an exploding superheavy while not vaporizing itself?

[@EagerSpace]

It's an interesting question. You can probably come up with scenarios where it won't work, but remember that in the AMOS-6 explosion on the pad, the payload survived both the second and first stages blowing up and appeared to be largely intact when it fell over soon after.

[@J7Handle]

@@EagerSpace I'd guess that before we had a hotstaging ring, a successful Starship pad abort was unlikely, but with it, it's now very possible. We know these engines can start up in around 1 second, if the ship were falling for that long, it would take another second or two to halt the fall and start ascending, at which point the ship has dropped up to 20 meters, not great. However, if the ship is also being held by the tower arms during fuel up, then maybe a pad abort is more likely. There are also pad abort scenarios where the abort could be triggered before any actual explosion occurs, for example, a rupture in plumbing dumping methane everywhere could justify a pad abort, albeit could also cause the explosion it's trying to escape from.

[@RoBear-xo6zw]

Crew Starship is not a rapid launch abort vehicle , like the Crew Dragon capsule is… Crew Starship, needs to be modified and have a modified Crew Dragon capsule on top, for takeoff and landing. Crew Starship lunar lander, needs the Crew Dragon capsule on top as well, for aborting lunar landings and take off… Just relocate the helium pressure tank and rebalance the setup 🎉 The modified Crew Dragon capsule, will use the same location for the abort thrusters and have a pass through portal hatch located in the center of the heat shield 😂🎉… for Starship ingress and egress 🎉🚀

[@ToaArcan]

I think my two main concerns when it comes to Starship aborts are: 1) Can Starship actually get away from a fatally-malfunctioning Superheavy before it explodes? Like in the scenario that LES is needed, the main purpose is to get the crew the fuck out of dodge before they die, and injuries from the sudden increase in G-forces could cause become a secondary concern. Hence why NASA uses solid rocket motors for their tractor rockets. Starship ain't the _quickest_ to get going. It's a liquid-fuelled rocket, that's to be expected, and it's huge. That's a lot of mass to shift in a suddenly-developing crisis where they have seconds to get away. 2) What if the fatal issue is with Starship itself? It's a whole-ass rocket stage unto itself, if it's the part that's going wrong, are the crew not just toast? This was an issue the Shuttle had too, and while with hindsight neither of the fatal shuttle incidents were "Orbiter malfunctions during launch", that was never guaranteed and the lack of a proper way to save the crew from the vehicle itself becoming the problem wasn't really acceptable. And sure, you can apply that question to capsule-based designs too, and they _have_ gone wrong in the past. Apollo 1, Soyuz 1, and Soyuz 11 can all be considered examples of such, as can Apollo 13, though that one was thankfully not fatal. But there's a whole lot less to go wrong with those, because they're not acting as stages.

[@EagerSpace]

1) The answer is "it depends on the scenario". Supposedly there's a "fast start mode" for Raptor that isn't kind to the engines but gets them to spin up faster. And Starship 3 will have a higher thrust/weight ratio and get away faster. I'll also note that booster failures are very rare and when they do happen they tend to start at the bottom of the rocket at the engines. Super heavy has a lot of engine redundancy and in most cases if an engine is starting to misbehave you just shut them down. 2) Depends on the scenario. Starship scenarios would mostly be reentry ones, and those depend on working aerodynamic surfaces and lighting the landing engines. The flaps have redundancy and you only need to light 1 out of 3 engines for landing, so plenty of redundancy there. I'll also note that if SpaceX flies starship as often as they say, they will have a ridiculous amount of empirical data before they put people on it. It will be far better understood than any existing human launcher.

[@andreyplumer3990]

It's interesting to see possible abort modes for Starship's lunar landing scheme. It looks like there are almost none at all which makes the whole scheme very fragile (and strange). On the other hand, why we can't use the Starship to deliver a really safe LEM to the moon? It feels like there's enough capability for that.

[@EagerSpace]

Starship has 6 engines that are enough to slow down and land on the earth, which means lunar starship will have a fair bit of engine redundancy (assuming main engine use; not clear what their thruster approach is as it hasn't been disclosed). The lunar module from Apollo had low redundancy because weight was so important - it used very simple engines but there failures in some cases were problematic. If you wanted to use a different lander, you would need something more capable than the apollo LEM, as the orbit for Artemis requires more energy than the one used for Apollo. It's certainly possible to do a vehicle more like the LEM - that's what the national team led by Blue Origin proposed - but SpaceX's bid using Starship was much much cheaper than the National team bid and NASA judged it to rank highly technically and to be the only one NASA could afford.

[@aco2518]

Is Raptor able to light without having completed an engine chill first? I'd think dumping cryogenics into a warm engine would be a nightmare no?

[@EagerSpace]

Interesting question. My *recollection* is that Musk talked about this once and said that Raptor has a "start right now" mode (not what he called it) that could be used in situations like this but that it would probably prevent reuse of the engine.

[@gasdive]

A crew Starship landing in water will fall over. That's a 50-70m fall. That's probably not survivable, particularly as the orientation of the forces is going to be random.

[@EagerSpace]

That will be fun to try to analyze... * The acceleration won't be G the whole way; at 45 degrees the height is already 30% less and the acceleration will only be about 0.7G * It will initially sink into the water some amount because of the mass of the engines, and that will both reduce the height and slow down the rotation * As it tilts over the water provides move force against the bottom to slow down the rotation * The circular shape will push quite a bit of water out of the way and that will absorb a lot of energy. You would need a fluid dynamic simulation to get a real answer, and I'd love to see one, but I don't expect to be at all like a fall from that height.

[@gasdive]

@@EagerSpace we may get some numbers on the next water landing. If the cameras survive and it's daylight. The people will be near the nose, which may fall faster than 1g at some points, even if the CoG is falling much slower than 1g. It might be like being in the bow of a ship in really extreme weather, or it might be really bad, I don't know.

[@jmstudios457]

Escape system acceleration really depends on the energy of the propellents and how fast the shockwave moves. Aerozine50/N2O4 on the Titan II launch vehicle were lower energy, and the acceleration of the ejection seats was more than enough. The reason why Orion's abort system has much higher acceleration is because its liquid hydrogen propellent is much more energetic in terms of hydrogen/oxygen. My main concern with Starship is pad and early in flight aborts, the RVACs will have much less thrust ASL. There is quite a lot of liquid methane and liquid oxygen in there, and I'm not sure if starship would have the oomph if needed to get away. It also depends on the ignition system of the engines. If it uses a spin start system, either a solid cartridge like the LR-87 and LR-91 family used on the Titan rockets, or a pressurized gas system, on the RS-68, they might be able to start the engines in time for abort, if they detect the problem early enough. If they use a Tank-head/bootstrap method used on the F-1 and RS-25, then it would be infeasible as these take considerably longer to start

[@dsdy1205]

If Elon's tweets are anything to go by, the Raptors can spin up very very quickly. The flip and landing at least, shows that Starship can create time critical ignition in a rapidly varying g-environment

[@EagerSpace]

Thanks. Musk did comment - which I interpret as a speculation - that it would be possible to spin up the Raptors quicker in an abort scenario. I don't think there's been a lot of engineering brainpower spent on the need to spin up motors quickly simply because it hasn't been a requirement AFAICT.

[@JC-IV]

Is there an update to this given hot staging?

[@EagerSpace]

Hot staging is clearly an enabler for starting up starship if there is a problem with the booster - just spin up the engines and go.

[@ryanrising2237]

Thanks. I think the pad abort section shows pretty well why Starship’s “use the main engines” option on the upper stage isn’t really a valid answer to the scenario, _even if_ we assume a compromised Superheavy doesn’t also damage the upper stage.

[@EagerSpace]

I'm not sure what your point is. Why isn't using starship's engines a valid answer?

[@ryanrising2237]

@@EagerSpace It’s just as you pointed out - the engines need to have time to spool up, and even if they can be brought online quickly enough in a disaster case they don’t provide anywhere near the kind of acceleration that you’d want to try to move yourself away from an exploding booster.

[@topsecret1837]

@@ryanrising2237 There’s also the acceleration from pneumatic pressure sending the ship upwards a short height if that section were suddenly to increase in pressure from the engines venting gas into that direction. With Raptor having a high chamber pressure I don’t see how that would be a problem in terms of high-> low pressure. Also during ascent the booster could use gridfins to increase its own drag to force itself to slow down away from the ship.

[@ryanrising2237]

@@topsecret1837 There are plenty of options for separating the two if everything is working properly, yeah, but it’s a good idea to design an abort system specifically to work when the rest of the system isn’t.

[@engineerstudentshenanigans3452]

@@topsecret1837 Pressure outside of the engine, even a distance away, is surprisingly important to the operation of a rocket engine. For Example, the Saturn I was extensively designed around aerodynamics of the "skirt", and different exhaust paths not creating any sort of interference with each other, as this could potentially send a shockwave up the plume and into the engine and destabilize it.

[@NikanDragosysSerpenDra]

LOL🤣LEVIATHAN OUR LORD.

[@Maskddingo]

Great videos. Very informative!