Slashdot Mirror
Columbia Accident Board Preliminary Recommendations
fwc writes "The Columbia Accident Investigation Board (CAIB) has released some preliminary recommendations to NASA - To do a better job at inspecting the leading edge of the shuttle's wings, and also to ensure that pictures of the orbiter are taken while in orbit. More recommendations are to follow in the full report which is expected in June. More detailed information on the recommendations are at space.com and spaceflightnow.com. NASA Administrator O'Keefe seems optimistic that they will be able to return the shuttle fleet to flight by the end of the year since there has been no show-stopping problems which have been discovered during the investigation."
Would you use software that crashed 1-in-50 times
;), I reboot maybe once every couple of weeks. If we said an average of once a week, we're talking one OS crash every year, now that's not too shabby. If we're talking web servers that crashed every 50th http request, that obviously would not be good. If we're talking web broswers that crashed every 50th page request, that would suck. If it crashed every 50th time I fired it up, that would be great (again since I have a usage pattern that starts the browser once and I never close it).
Depends a lot on the software and what you mean by 1-50 times. As an example, take your OS (please
The shuttle is similar, given that almost any problem can easily turn into a catastophic problem, how much of that failure rate is intrinisic in the activity (e.g. no matter how safe you try to make mountain climbing, there is always an element of risk that is higher than many other activities). And the frequency of that activity, if we're talking 50 missions at two missions a year, that's a lot of years between failures. Hey, that's what makes being an astronaut what it is, a risk, that's why they are elevated to such a high status (unfortunately often times not until AFTER something bad happens).
Honestly, do you have any contingency to examine in space and fix the shuttle if it does have problems? No, well, see you back here in another 10 years.
Exactly, especially when there are rockets or rocket engines involved. Rockets work via a large controlled explosion. The larger the explosion (the more thrust), the harder it is to control that explosion. Anytime you're strapping people into a vehicle that has close to 6 million pounds of thrust behind it, you're taking a risk that the explosive power behind that ~6 million pounds isn't going to get away from you. And it doesn't take a rocket scientist to see that. ;)
My journal has hot
You are at 30,000 feet in your 747. You've been flying for 11 hours. You decide to go and do a prelanding check because hell, anything could have broken and you've got 200 innocent people on board. In your checklist you discover that one of the wings has lost 14 of its 18 welding points.
You can't repair anything away from your repair facility. You can't land the thing any differently than you normally would to reduce stress. And you can't transfer your passengers to a different plane while in the sky. There's no parachutes. Why did you even bother checking?
And that's a 747 very close to the surface going much slower built with much less exotic materials.
Both Challenger nor Columbia were caused by human error. In Challenger's case, the politicians/managers made the decision to go despite warnings from the engineers. In Columbia's case, they had the opportunity to take pictures of the shuttle in orbit, per suggestions by the engineers, but decided not to do so. (What they could have done to save the crew is a separate topic.)
So when we talk about the dangers of space flight, or how unreliable the shuttle fleet is, let's not forget how much of an element human decision-making is.
-Thomas
when the shuttle launched, a piece of debris broke off and hit the wing. Back then they said it didn't matter, then the shuttle exploded on re-entry. Now, months and months of 'careful study' they find that the wing had been damaged. No sh*t... what a useless exercise. And the recommendation: study the shuttle more carefully! Ummm. yeah, how much are they being paid for this?
A classic misinterpretation of an accident report - though this isn't even a full accident report yet, and I imagine there will be even more misinterpretation when it is finally released.
What the investigators have actually determined is really nothing. What they have determined probably happened is that there was pre-existing corrosion to the frame of the wing's leading edge, which weakened it to the point where the foam strike caused something to break. This pre-existing corrosion should have been caught and fixed by NASA, and if finally proven as fact, would be the root cause of the accident. The foam hit was not the cause of the accident, the corrosion was. Assuming they stick to this theory, of course.
I've said before that almost all accidents are a series of events, some preventable, some not, most benign by themselves. It's that particularly series of events and the way they unfold that causes the accident. Without the corrosion, the foam hit would have done nothing. It's happened so many times before without incident, and the shuttles were built to take punishment - these are vehicles designed for repeated launch and re-entry, for God's sake - the G-forces, shock and vibration they're built to withstand are almost ridiculous, and they've been hit by multiple objects at launch, in orbit and during re-entry before without incident. The facts seem to suggest that Columbia was no longer in like-new condition - that it was fatally weakened even before its last launch. If it wasn't for this foam hit, it would have been something else that would have brought it down eventually. The foam was just a catalyst.
What I find shocking is the apparent deriliction of maintenance on the part of NASA, and the budget cuts really need to be looked at as a contributing factor to the accident. There's no way these shuttles should be allowed to have this kind of corrosion, and Columbia was just refitted a couple of years ago - the wings were taken completely apart, they should have seen any damage like this. Even if they didn't, though, they should be doing MRI's or whatever they need to every 6 months or a year to check the interior structures of all critical structures.
Just one final comment - someone suggested doing 2 "pre-flight" checks, one before launch and one before re-entry. This doesn't make any sense whatsoever. The poster used commercial airliners as an example - well, this would be like doing a "pre-flight" check both before takeoff and before landing. First of all, the pre-flight on a commercial airliner is usually nothing more than a walk-around by the pilot and a systems check while taxiing (many airplanes spend 30 minutes or less at the gate before pushback). The space shuttle sits in a hangar for 6 months being looked over with a fine tooth comb before launch - it's much more thorough than anything a commercial airliner goes through. Second, there's no "pre-flight" before a plane lands - that would not be feasible or even necessary. There's no reason why a space shuttle would need such a check either if the vehicle itself is in good working condition - which should be established while it's on the ground, not in space. If you establish the fact that the foam hitting the wing was not catastrophic in and of itself but that it was corrosion to the interior structure of the wing's leading edge that weakened it and led to the break when the foam hit - that's something that should be caught before it even gets to the launch pad. It's not something you should worry about in flight.
This begins to address the real problem. The space shuttle was sold as "routine access to space." It isn't. It's a routinely operated experimental vehicle. That's not good. Back in the 1940s we didn't build Bell X-1s for the Air Force. We used what we learned from the X-1 to build production jet aircraft.
Official attempts to build better rockets (NASP, X-33) have failed to produce even flyable vehicles. Currently a considerable number of people have given up hope that the aerospace establishment will eventually come up with a vehicle that actually gives us routine access to space. I believe Larry's friend and coauthor Jerry Pournelle is one of them.
People have noted that real innovation in software comes from academia and small companies. Microsoft talks about innovation, but doesn't really deliver.
In the aerospace field, however, a healthy culture of small companies and independent academic research hasn't begun to exist until recently. And NASA's experiments turn into expensive failures. What's worse is the establishment tends to inadvertently suppress research by people other than itself.
"Beer is proof God loves us and wants us to be happy." -- B. Franklin
The space shuttles are man made vehicles designed to take people into space! There are going to be inherent risks with such undertakings, but this is the nature of space exploration. Time will provide safer alternatives, but for now 1/50 isn't bad.
Really? The Mercury/Gemini/Apollo program didn't kill anyone in a flight (3 were killed on the ground and another 3 came about as close as possible) and that was in the 60s and they were going to the moon. The reason the space shuttle has a higher failure rate is simply that it has more moving parts and things to go wrong. The shuttle failure rate would be significantly higher if it really flew once a week as it was designed to and if the per flight costs were what they were expected to be. Doesn't the fact that it flies 1/50th of the amount it was designed to tell you something about the difference between the expected failure rate and the actual failure rate?
The astronauts know these risks too, and they willingly assume them.
They are brave people, no question. I'm sure, given the choice, they would rather fly in a safer space craft and risk there lives for something more important than studying the effects of weightlessness on tiny screws.
And what if the columbia had broken up over a populated area of California rather than empty portion of Texas. Would all those people who gave their lives appreciate the risk that was being taken on their behalf?
PS: The Internet Explorer comment is unnecessary.
Well IE never killed anyone (although I could be wrong on that) -- they are both crap though.
I am not a number! I am a man! And don't you
I really don't think it's a hallmark of poor design.
The orbiting capsules were intended to go into orbit for a matter of hours and then come back to earth. They were not reusable and were far too small to contain many science experiments. The shuttle is intended to go into space for extended periods of time. This requires more equipment, more moving parts. It is also intended to be a scientific laboratory. This, too, requires WAY more equipment, and a lot more moving parts. It's also intended to be reusable.
This isn't a poor design/good design issue. It's goal oriented issue. The capsules were considerably simpler, because the goals of the missions were considerably simpler. The shuttle is more complex (read: has more moving parts) because the goals of the shuttle missions are more complex.
You may have an argument for the Apollo missions: more complex missions. But Apollo 13 was almost a disaster, and many people in the field consider it a miracle that one of the Apollo missions didn't go wrong.
The shuttle missions can rarely be compared to the early-NASA missions. It was a different world, there were different goals, a different government, and different public support. Yes the missions happen less often than they were originally intended, but then again, there's far less public support of space missions, and Congress cuts NASA's budget practically every year. What do you expect?
The best way to predict the future is to invent it.