Slashdot Mirror
A Look At Orion's Launch Abort System
An anonymous reader writes: With the construction of Orion, NASA's new manned spacecraft, comes the creation of a new Launch Abort System — the part of the vehicle that will get future astronauts back to Earth safely if there's a problem at launch. The Planetary Society's Jason Davis describes it: "When Orion reaches the apex of its abort flight, it is allowed to make its 180-degree flip. The capsule of astronauts, who have already realized they will not go to space today, experience a brief moment of weightlessness before the capsule starts falling back to Earth, heat shield down. The jettison motor fires, pulling the LAS away from Orion. ... Orion, meanwhile, sheds its Forward Bay Cover, a ring at the top of the capsule protecting the parachutes. Two drogue chutes deploy, stabilizing the wobbling capsule. The drogues pull out Orion's three main chutes, no doubt eliciting a sigh of relief from the spacecraft's occupants."
Possibly. Probably not.
The failure modes for the Shuttle are unlike any other spacecraft's - even the near-clone of it, Buran. And any theoretical abort mode for it has to account for that weirdness.
First, the Shuttle has to remain intact. You can't just eject the "pilot area", because the whole thing is really monolithic. You might be able to get away with ejection seats, but that works only for a very small period of spaceflight (probably not Challenger - they'd have ejected into a fireball and coasted up to 60,000ft). They did, in fact, have some ejection seats on the early test flights, with partial crews, but they did away with them in use (letting some escape while leaving others to die was inhumane, and making all seats eject was far too heavy for the marginal benefit).
Second, the boosters cannot be shut off. That's the big safety drawback of solid rockets - you light them, and they aren't going out until they're out of fuel. This means detaching the boosters isn't going to work, because (without the drag and mass of the Shuttle holding them back) they'll just blow past the Shuttle, bathing it in hot exhaust. If my memory is correct, the Shuttle is the only manned rocket in history to use solid engines, in no small part because of this sort of problem. Even the Soviet shuttle clone, Buran, used all-liquid engines.
Third, the Main engines are nearly useless in-atmosphere. They're lit mainly because they sometimes fail to light, and having that failure occur halfway to orbit would suck. The "boosters" provide about 80% of the thrust, if memory serves. The SSMEs aren't even at full throttle for much of the flight - Challenger had just set them to full when the stack exploded. So any idea of "just floor the main engines to outrun the boosters" is ludicrous.
Fourth, these sorts of disasters happen with very little notice. Rocket fuels are generally extremely volatile - even the least exotic combo, LOX+RP1, is still liquid oxygen and high-grade kerosene. LH2 is safer than some things (ClF3 was, and still is, considered for rocket use), but it's still pretty dangerous, and when a tank of LH2 and LOX decides to explode, it's not going to give you even a second's warning. So the escape systems they did add after Challenger probably wouldn't have been usable, because it literally involved jumping out of the Shuttle.
Fifth, the Shuttle is HEAVY. Really goddamn heavy, especially since you're not going to be able to dump the payload during an abort. So you've got the crew, all their supplies, whatever they were carrying to orbit, and all the vehicle mass. Any rocket that could accelerate the Shuttle away from an exploding stack would be itself enormous, not something you could really justify launching into orbit every mission.
Because of these peculiarities, the Shuttle abort modes are along the lines of "pick where to crash" instead of "run away from the explosion". The four post-launch modes are "return to launch site", "trans-atlantic landing", "abort to once-around" and "abort to orbit" - all of which require a mostly-working Shuttle and must be used after the boosters are exhausted.
An LES like this could not have saved them, because you couldn't really use an LES such as this on the Shuttle. Modifying the Shuttle enough that an LES like this makes sense would basically require making it not a Shuttle - in fact, you'd basically end up with an Orion-like capsule on top of an SLS-like stack, because they're literally reusing that much of the technology.
*sigh* This is one of the biggest pieces of misinformation about solid rockets floating about out there, spread and repeated by shuttle detractors in a cargo cult like fashion until it's now regarded as a law of nature. What most people (including engineers who should know better) don't realize is that you don't need to shut them down in the first place- you just need them to produce net zero thrust. This is done via blowout panels in the front dome, and sometimes by blowing off the nozzle as well. And it's not like this is a new fangled technique either... It was used on the Polaris A-1 and A-2, Poseidon C-3, SUBROC, ASROC, Minuteman I and -II, and Peacekeeper missiles. It would have been used of the SRB's of the Titan IIIC booster for manned Dyna-Soar and MOL launches. It's used by Minuteman III missiles...
It wasn't used by the Shuttle because during the SRB burn, the SRB's are essentially 'dragging' the ET behind it... and thrust termination would have resulted in them 'hanging' from the ET or having to be jettisoned and the resulting changes in structural loads would have shredded the ET and tossed the Orbiter into the airstream where it would be broken up. (Which is essentially what happened to Challenger.) A normal SRB jettison doesn't shred the ET, because the loads come off gradually as SRB thrust decays and they're jettisoned as the T/W ratio passes through 1.
NASA looked at using an Orbiter mounted solid rocket to power it away from the stack, but even if the motor was used on a normal flight for orbital insertion after ET jettison it was too heavy.
A friend of mine, an aerospace engineer by trade, once explained it thusly - "during first stage flight, the SRB's lift the ET and the SSME's lift the orbiter". This isn't entirely true, but it's a useful first approximation. And that being said, other than a brief time right around Max-Q (when the throttles are backed off to control aerodynamic loads) and as MECO approaches (when the throttles are backed off to control G loads) the engines are in fact run at full throttle during powered flight.
If only the decision was that simple... Sadly, it wasn't.
First there were performance issues; The solid motors need to match to within 5% of each other - which proved essentially impossible to achieve with a monolithic grain as the propellant tended to stratify during the extended pour and the extended curing time. The solid motors needed to have consistent and predictable performance during the burn - which was almost impossible to achieve due to the aforementioned stratification problems. Both problems were also made worse because they couldn't figure out how to safely mix and pour the grains for both boosters in a single batch. Segmented grains, which could be poured in LH and RH segments from a single (smaller) batch suffered from none of these problems.
Next, there's storage and handling problems. The larger the grain, the heavier it is, and the harder it is to prevent it from flowing and deforming under it's own weight. Equally, since the large grains have to be cast upside down they have to be rotated rightside up - and nobody knew how to do that with large monolithic grains. A flex of as little as a couple of millimeters could crack the grain or lead to delamination. Also, segments could be stored individually, reducing fire and explosion risk.
Inspecting the grains with the technology of the time was also several orders of magnitude harder for a large monolithic grain.
Lastly, while there was a only a limited base of flight experience with large segmented grains (via the Titan IIIC)... there was no flight experience with large monolithic grains.
tl;dr version - there were a lot fewer known unknowns with segmented solids than with monolithic solids. A number of the known unknowns for monolithic grains were either outright show stoppers or could result in ruinously expensive R&D programs to discover if a solution was even possible. The known unknowns for segmented grains were all issues of scaling from existing experience.