Personal observations on the reliability of the Shuttle (by Richard P. Feynman)

Virtual School ^ | ? | Richard P. Feynman

[js1138]

Peter Schickely, for those not tuned to the world, is the famous musicologist fro the University of Southern North Dakota at Hoople, who discovered the unknown composer, P.D.Q. Bach.

[monkey]

It's not a stationary process. After a failure they presumably learn something, and are less likely to fail in the same way again. This is true of a catastrophic failure, or a minor failure that doesn't affect the mission. Note that by calculating the odds right after a failure, your method is highly biased.

A better way to calculate (even rough) odds would be by doing time between failure analysis, i.e., looking at how many flights until the first failure, and how many flights between the first and second failure, etc.

[Poohbah]

The computer system is very elaborate, having over 250,000 lines of code.

Ah, such a simpler era...

250K lines of code probably wouldn't make a decent spreadsheet program these days.

[185JHP]

They get more than money - they get paid to do what they enjoy doing. Kinda like LEOs - "money for nothin', and the kicks for free!"

[boris]

Not to mention the Chicks.

[KellyAdmirer]

Your points are well taken, but I think 2% is a fairly realistic estimate. You have to pick your endpoints somewhere. There are many ways to figure the odds. They give varying results but in my opinion fall around the same area.

1. Suppose we did the odds of death from the shuttle after the first 50 shuttle flights. 1 failure out of 50 = 2%. If we did it after 100 flights, sure, it would be 1%, but we already have the second failure after only 13 flights into the second 100-flight cohort, so the shuttle better fly the next 87 missions without fatality or that percentage goes back up. Since it has been about 80 flights between crashes, that may or may not happen.

2. We could calculate it based on persons who have died as a percentage of those who have flown. This, in fact, might be the most useful statistic. Since Yuri Gagarin in 1961, approximately 400 different humans have attempted to fly into space. At least 17 died (I say "at least" because there are suspicions of unreported Cosmonaut deaths). 17 out of 400 is about 4%. That percentage would be higher if it were just US astronauts, btw. I saw an astronaut quote this figure on tv today, so it is not something I pulled out of the air but rather the kind of figure real-life prospective astronauts think about, however slightly, when signing that release.

3. We could look at the average time between major catastrophes. The three years are 1967, 1986 and 2003. A nineteen-year interval followed by a seventeen-year interval. While there were a lot more flights in the latter period, we also were much more technologically sophisticated and had learned from a lot of mistakes. That suggests that the odds of dieing as a function of time are staying about the same or even increasing.

My opinion from the facts is that boarding a government spacecraft for a spaceflight has a statistically significant death rate in the low single digits, around 2% and maybe higher. As long as the technologically obsolete Space Shuttle keeps flying, I doubt those odds will go down much no matter how many lessons NASA learns from each crash. I would make sure my life insurance were fully funded before boarding one of those ageing deathtraps.

[HighWheeler]

here you go:

I have argued flying saucers with lots of people. I was interested in this: they keep arguing that it is possible. And that's true. It is possible. They do not appreciate that the problem is not to demonstrate whether it's possible or not but whether it's going on or not.

— Richard Feynman



I think that it is much more likely that the reports of flying saucers are the results of the known irrational characteristics of terrestrial intelligence than of the unknown rational efforts of extra-terrestrial intelligence.

— Richard Feynman

[Quix]

Thanks much.

I think I've read something else as well. . . something in the other direction. But it may not have been first hand. Will try and search myself.

MUCH APPRECIATE YOUR HELP.

BLESSINGS,

[Quix]

cute.

interesting.

THANKS.

[Quix]

THANKS.

WILL CHECK THE LINK OUT.

MUCH APPRECIATED.

[theFIRMbss]

>all 13 of his examples have been solved

One thing to look at
will be the thirteen "fixes."
Bugs feed on kludges...

[theFIRMbss]

>250K lines of code probably wouldn't make a decent spreadsheet program these days.

Yes, well, remember
one of the big arguments
for a big OS

integrated to
a graphic interface was
that all the shared code

would allow user
programs to get much smaller!
(If we all used FORTH...)

[boris]

"One thing to look at will be the thirteen "fixes." Bugs feed on kludges..."

As if you understood a word. These were SSME (main ENGINE) problems. Just FYI, the engines are TURNED OFF during descent.

Oy.

--Boris

[theFIRMbss]

>As if you understood a word. These were SSME (main ENGINE) problems. Just FYI, the engines are TURNED OFF during descent. Oy.

Hey, take a chill pill.
Just about every system
has been fiddled with

since the Challenger.
All those "fiddles" could add up
to lots of strange bugs.

And they probably
burn the main engine to slow
their orbit and start

re-entry. (Save your
oys and the duh FYI's
for people who care.)

[boris]

"And they probably
burn the main engine to slow
their orbit and start"

They do not. There is a separate propulsion system called the OMS (Orbital Maneuvering System) which does the retro burn. It is as different from the SSME as a snail is from a lion.

--Boris

[theFIRMbss]

>They do not. There is a separate propulsion system called the OMS (Orbital Maneuvering System)...

"The OMS engines burn monomethyl hydrazine fuel (CH3NHNH2) and nitrogen tetroxide oxidizer (N2O4). Interestingly, when these two substances come in contact, they ignite and burn automatically (i.e., no spark required) in the absence of oxygen. The fuel and oxidizer are kept in separate tanks, each pressurized by helium. The helium is used to push the fluids through the fuel lines (i.e., no mechanical pump required). In each fuel line, there are two spring-loaded solenoid valves that close the lines. Pressurized nitrogen gas, from a small tank located near the engine, is used to open the valves and allow the fuel and oxidizer to flow into the combustion chamber of the engine. When the engines are shut off, the nitrogen goes from the valves into the fuel lines momentarily to flush the lines of any remaining fuel and oxidizer; this purge of the line prevents any unwanted explosions. During a single flight, there is enough nitrogen to open the valves and purge the lines 10 times!"

-----------------------------------------------------------------------------------

There you go! You see--
it's nicer to be helpful
than a smart aleck...