SpaceX’s Pad Abort Test | Hacker News Reader

SpaceX’s Pad Abort Test

spacex.com

162 pointsggonweb68 comments

At the Kennedy Space Center, NASA had an interesting abort procedure for astronauts still on the ground at the launch site: an underground Rubber Room which astronauts could slide down into.

[0] https://scriptunasimages.wordpress.com/2012/11/23/inside-nas...

[1] http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/2011001...

And to contrast this with the Orion capsule abort system: Crew Dragon's abort thrusters will eventually used for landing, while Orion's (much heavier) abort system is jettisoned 120 seconds into flight.

See https://en.wikipedia.org/wiki/Orion_%28spacecraft%29#Launch_... and https://en.wikipedia.org/wiki/Orion_abort_modes

Even if they are not used for landing, main thing is that they use same fuel tanks as those used by Draco thrusters for orbital propulsion and deorbit, so no weight is wasted (weight of the Super Draco engines themselves is not that big). Overall payload loss for the abort system is about same as if using the escape tower (ca. 800kg if i do my math right), but escape burn is possible throughout the ascent, not just in the first 3 minutes, and there is no additional safety risk from it (if escape tower failed to separate, astronauts will be doomed, here there is no such risk).

I know the SuperDraco's are designed to eventually enable a soft-landing back on Earth. In this case they are only being used for the abort, and the landing will be by Parachute, presumably the powered landing approach isn't ready for testing yet. I understand the initial manned launches will use parachutes for landing as well.

What I'm wondering is, will they eventually carry enough fuel to both be able to do an abort, and then do a powered landing, if required? I assume so, since at that point the only cost in terms of weight to be able to do the abort as well as the landing would be the weight of the fuel for the abort itself.

Also if the capsule will be capable of an abort at any stage in the ascent, what preparations will be in place to allow for capsule recovery? Presumably it could come down anywhere along the line of flight, up to the point at which it achieves orbit. That could include the middle of Africa, and maybe even the Indian Ocean. Imagine a second stage engine failure.

None of these are criticisms of the system, it's fantastic engineering and potentially a huge win over current options. Excuse me while I got and watch Gravity again (while switching off the part of my brain that knows from KSP how orbital rendezvous work).

>What I'm wondering is, will they eventually carry enough fuel to both be able to do an abort, and then do a powered landing, if required?

Even if it could, its important to note that the abort system would only be used when something has gone drastically wrong, potentially an explosion, which means you can't garuntee Dragon will be undamaged. In that case, would you trust a complex powered landing, or a simpler parachute landing where you have triple redundancy?

In addition, SpaceX launch on the coast and the rocket flies over the sea, so an abort may end up with a sea landing. I haven't seen any concepts from SpaceX showing a powered dragon landing at sea, and it would presumably require a pretty calm sea state, while parachutes would work in any weather. You could see similar problems with a powered landing on unfamiliar land, for instance trying to "land" on a tree canopy.

Why would parachutes work better at sea? Just like a rocket, the parachute stops providing useful upwards force as soon as the capsule touches water/ground. The only thing I can think of is that it is probably easier to add landing air bags to a capsule descending by parachute than by rocket, but that's really only useful on land.

Parachutes work better at sea than land; water becomes far softer as you reduce speeds. Also, a powered landing on a 30 foot wave seems like a rather complex problem not to mention buildings etc. Worse, powered landings have little redundancy where a sea landing with 2/3 of your parachutes is not a major issue.

I suspect a larger issue is seawater is not something you want anywhere close to delicate equipment. But, for an emergency landing that's not much of a concern. Anyway, I suspect these are designed so if either the powered landing or the parachutes work the landing is survivable even if the craft is damaged.

The super-dracos are on the sides of the capsule, not the bottom, so should be able to provide thrust all the way down to surface contact on land and sea. Also the vehicle can suffer the loss of a thruster without compromising it's abort or landing capability.

I don't really see that parachutes in 30ft waves are going to be all that much beter than anything else, but don't manned launches have fairly strict weather condition safety requirements? I'm pretty sure that would include sea states in any conceivable abort landing zone.

One final advantage is that the vehicle can use the thrusters to manoeuver prior to landing, potentially avoiding problematic terrain. In theory you could probably designate a series of preferred abort destinations that are within the vehicle's operational envelope for different stages of the flight. So instead of ending up anywhere along a line, there would be a series of preferred landing zones along or near the line.

Thanks for the good point about manoeuverability when landing on bad terrain during an unplanned emergency.

> don't manned launches have fairly strict weather condition safety requirements? I'm pretty sure that would include sea states in any conceivable abort landing zone.

You might be right, but note that all the shuttle abort landings sites were on land at a handful of locations, in part because the shuttle was a glider (so it had a decent amount of unpowered range) and in part because landing the shuttle somewhere that wasn't planned was nearly impossible (since it needed such a huge flat expanse to land). Therefore, you only needed to have a good weather simultaneously at a few locations.

For the Dragon, however, the lack of gliding ability and the ability to land anywhere in an emergency makes it infeasible to wait for simultaneously good weather everywhere, simply because it might end up anywhere along it's flight path in an emergency. So you probably need to be able to handle bad weather.

> Parachutes work better at sea than land; water becomes far softer as you reduce speeds.

Sorry, my comment was ambiguous. I meant "Why would parachutes work better than rockets at sea?" not "Why would parachutes work better at sea than land?"

> a powered landing on a 30 foot wave seems like a rather complex problem

This is only an issue if you're greatly decelerating in the last 30 feet, since then that window is shifting up and down with the ocean. However, I'd guess that all the major deceleration has already happened by this point, and the rockets are just maintaining a slow steady descent speed for the last 30 (or 100) feet. (Would greatly appreciate anyone who can tell me I'm wrong here.) In this case, they function just like a parachute.

> Worse, powered landings have little redundancy where a sea landing with 2/3 of your parachutes is not a major issue.

As simonh mentions, the capsule has redundant thrusters, which again seems very analogous to a parachute.

> Buster the Dummy already works for a great show you may have heard of called MythBusters. Our dummy prefers to remain anonymous for the time being.

I'd love to know what the SpaceX test dummy's name is, but for now I'm going to assume it's Edison.

My prior has a prominence near "Elon".

I'm betting on Chuck (as in Yaeger)

Unlike MythBusters, SpaceX is likely to have more than one dummy and, unlike Nasa^H^H^H^HESA they tend not to make cutesy circus out of their projects, e.g. avoid things like "Oh, hai, Twitter, I'm the Rosetta probe! Yuppie!"

So I suspect dummies will remain anonymous.

I don't think you're the target demo for the cutesy circus events. Unlike SpaceX, NASA needs all the public exposure and goodwill it can get, which is what the Twitter accounts are for. There are plenty of data and science-heavy websites for the non-casual user.

That's exactly my point. SpaceX doesn't need to name the dummies, nor judging by their history they would want to.

Just Read The Instructions

Of Course I Still Love You

Fair enough. I concede the defeat.

Omnia vincit amor et nos cedamus amori.

That's a quite myopic view of both NASA and SpaceX. The cutesy circus is pretty much a necessary condition of scientific popularisation-- NASA's social media experiments have been incredibly successful ways of getting the public to engage in space travel and science in a way that wasn't ever prevalent before.

It's trite, but also mandatory. I wouldn't be surprised if SpaceX didn't take some things tongue-in-cheek too; a war against humour is one nobody wants to fight. Except Boeing executives.

The trouble with wars against humour, is they tend to be really funny.

BTW, the cute Rosetta was from ESA PR, not NASA.

...indeed it was/is and I knew that. Thanks.

Did you know the Dragon capsule was named after the song "Puff The Magic Dragon"?

edit - and a wheel of Le Brouère cheese was on the first Dragon flight in honour of the Monty Python Cheese Shop sketch.

SpaceX named their recovery sea vessels 'Just Read The Instructions' and 'Of Course I Still Love You', after the Iain M Banks sci-fi Culture ships.

I wonder which of their vessels will win GCU Grey Area for its name.

If Elon launches a space vessel named 'Grey Area', I either want to be on it, or in a bunker.

> unlike Nasa^H^H^H^HESA they tend not to make cutesy circus out of their projects

They dress the dummy up like a cowboy.

http://www.wired.com/images_blogs/autopia/2012/12/A-2IXbMCEA...

Capt. "Ace" Owens?

It would be bad communication to give it a name if the vessel explode.

I'm assuming it's Jebediah Kerman

I agree, I totally expect that dummy to be named Jeb.

Thanks for this post. It makes clear this 'pad abort test' is really a test of the amazing SuperDracos which could constitute their own launch system! SuperDracos are partially 3D metal printed and can fire months after being fueled.

http://en.wikipedia.org/wiki/SuperDraco

Not just partly: from that Wikipedia article:

"It was announced in May 2014 that the flight-qualified version of the SuperDraco engine is fully printed, and is the first fully printed rocket engine. In particular, the engine combustion chamber is printed of Inconel, an alloy of nickel and iron, using a process of direct metal laser sintering, and operates at a chamber pressure 6,900 kilopascals (1,000 psi) at a very high temperature. The engines are contained in a printed protective nacelle to prevent fault propagation in the event of an engine failure."

The SuperDracos have already been through a full-duration simulation of the abort profile at SpaceX's test facility in Texas. It's more a test of system integration: that the guidance system is able to command them correctly, that nothing about the abort profile screws up parachute deployment, etc.

Wow what a ride! The capsule 'rocketed' off the pad, using only the superdraco thrust.

> Crew Dragon will accelerate from 0 to nearly 100 mph in one second. The entire test is less than two minutes long, with Dragon traveling over one mile in the first 20 seconds alone.

That sounds totally insane, and yet at 4.5 g's was what the Apollo crews experienced on a successful launch.

And 7 g's on re-entry ...

If my mental math comes out correct, that's 25 m/s in one second, so assuming straight against gravity, that's a bit over 3g for one second. Not too bad, considering how extreme of a situation is involved.

100MPH is about 45m/s, not 25. Fighting gravity, that's 55m/s^s acceleration, or 5.5g.

Still fairly mild, considering. The Soyuz launch escape system, for example, subjects its occupants to 14-17g for five seconds. Incidentally this is the only LES that's seen actual use, when it saved a crew from an exploding rocket in 1983.

You are very correct.

Whilst it is true that there has only been one launch abort so far, if the Shuttle had a launch abort system it would have saved the seven astronauts aboard. Just mentioning this in case someone mistakes your post as arguing that launch abort systems are rarely useful.

could, not would. Having a launch abort system is no guarantee that everybody will survive the ordeal.

That's a good point to raise. I didn't intend to make that statement but I can see how it could sound that way.

On the ground, drag racing achieves similar levels of acceleration too:

http://en.wikipedia.org/wiki/Top_Fuel

I think the difference will be the Apollo crews were expecting it and it was sustained and constant.

A launch abort isn't something that you can really brace for.

Everyone aboard will be strapped down in a nice flight couch. It probably won't be pleasant, but will be easily survivable--which sitting on an exploding rocket, the alternative, would not be.

That's actually fairly mild, as these things go.

The Apollo lunar missions experienced max G forces of 6.5-7.2g during reentry. (The Earth orbital missions only went to about 3.3g.) See: http://history.nasa.gov/SP-368/s2ch5.htm

Alan Shepard freaking flew Mercury 3 by hand during a 11.6g reentry.

Ejections seats are 12-14g or worse.

Of course, max G is only part of the equation; duration is the other. You can handle really high forces if they aren't for long. Given the short duration and lack of need to keep the occupants awake they could go a lot harder. I'm actually a little surprised it's so gentle.

Formula One drivers usually experience 5 g while braking, 2 g while accelerating, and 4 to 6 g while cornering.

One of the Soyuz (http://en.wikipedia.org/wiki/Soyuz_7K-T_No.39) hit 21.3g on re-entry.

Mercury escape tower used when booster loses control, 1961.[1] Unmanned launch with a dummy.

[1] https://www.youtube.com/watch?v=Vp9BnBDKa0s

54 years ago!!

Makes you think. What amazing feats of engineering. The abort capsule was even picked up by helicopter. Themselves relatively new at the time.

Apollo's launch escape system was tested using a rocket called Little Joe II, which started spiraling unexpectedly due to a gyro problem causing the rocket to fail. Fortunately the escape system worked as intended. [1] https://youtu.be/AqeJzItldSQ

I wonder how many of those observing realized it wasn't part of the scheduled test. What a lucky disaster!

It was also used with live people in a Soyuz rocket. Google up Vladimir Titov and there's a video of this. He attributed to the designer of LES for saving his and his compatriot's life.

As a general comment, I love watching SpaceX (for several reasons). The almost monthly stream of announcements, flights and "firsts" reminds me of watching the Apollo program progress as a young child. Then I was limited to seeing what NASA happened to put on television since I couldn't read for the early parts of the program. I'm finding reading briefs like this much more exciting as an engineer.

SpaceX's first manned flight is going to be a momentous occasion.

Nice explanation of what's going on. I've always thought there was an opportunity for much more of this kind of thing during the previous launches--explaining the process to those of us who really have no idea.

I know SpaceX is looking to make space flight more like commercial flight. Given that commercial airliners don't have anything like this in their design, shouldn't the goal be to get space flight to the same failure rate as commercial flight? Thus eliminating the need for an abort system?

Once they are that reliable, I'm sure SpaceX could consider removing the abort system to save on weight. Though I assume that the SuperDracos are designed to also be used in space / on Mars.

And for landing back on Earth, just not initially:

http://www.spacex.com/news/2014/05/30/dragon-v2-spacexs-next...

These thrusters would be there anyway. The fact that they can be used as the abort mechanism is just a bonus (a well planned and engineered bonus...)

The NASA Commercial Crew program's requirements include a line item for "assured crew return in the event of an emergency."[0] Which means any of the CCDev proposals that use the traditional rocket launch approach for the first stage need an abort mechanism that can carry the crew module away from the rest of the vehicle.

So arguably you're right, but they have to develop abort capability per the NASA spec.

[0] http://en.wikipedia.org/wiki/Commercial_Crew_Development#Req...

Airliners can't really explode, since they pull in oxidizer from the air rather than storing it on board. Since a fire will spread relatively slowly (relative to a rocket whose oxidizer and fuel decide to make friends all at once), the function of getting the squishy meat payload away from the fire can be accomplished through simpler means, in this case with inflatable slides and the squishy meat payload's own locomotive appendages.

That is to say, airliners do have an equivalent system in the form of their evacuation slides, it just looks quite different because airliners operate quite differently.

The history of escape capsules in military aircraft is quite fascinating, but not compatible with an airliner carrying 100s of passengers.

http://en.wikipedia.org/wiki/Escape_crew_capsule

Sure, but they're not there yet, and probably won't be for decades and several more generations of designs. And a much higher flight rate: remember, this is an industry where no system has more than 1000 flights, and most don't have 100, and all those are with vehicles that were destroyed in the process.

In the meantime, having an abort system is probably a good idea. And besides, it's a customer requirement. And you know how customers are.

I think the key phrase is 'more like' rather than 'the same as'. I don't imagine space travel is ever going to be the same as airline travel, it's always going to involve energies and therefore risks far in excess of commercial airliners.

Take engine failure. An airliner with completely failed engines and no motive power at least has a chance of gliding to a landing, without any additional safety or backup system. Rockets, not so much. Jet aircraft can recover safely from a whole range of technical issues, but while rockets have a certain latitude for recovery (SpaceX had a first stage engine failure during launch a while back but the flight continued) their resiliency is much more limited. I don't see that changing, and I doubt SpaceX does either.

> Thus eliminating the need for an abort system?

SpaceX's abort system is also its normal landing system. They'll have it available even if they never need it.

Reply on news.ycombinator.com