Post by john larkinPost by bitrexPost by bitrex" In retelling how the decision unfolded through the eyes of the
managers and the engineers, Vaughan uncovers an incremental descent into
poor judgment, supported by a culture of high-risk technology. She
reveals how and why NASA insiders, when repeatedly faced with evidence
that something was wrong, normalized the deviance so that it became
acceptable to them."
I guess I'm not grasping from the summary of the Vaughan book how its
conclusions greatly differ from the conclusions of Feynman et al.
Maybe if you read the book, youll understand. The conclusions could hardly
be more different, given the basic facts of the case. Boisjoly et al. and
the Rogers commission are only a fraction of the story.
Cheers
Phil Hobbs
I was able to at least find a summary, 620 pages about a disaster I'm
barely old enough to remember is a tall ask at this time.
<https://web.mit.edu/esd.83/www/notebook/The%20Challenger%20Launch%20Decision_1.pdf>
I think I somewhat understand the thrust of the argument, that nobody in
management really believed themselves to be taking risks of the kind the
public later perceived them to have been taking.
There was no particular person who was actively like "Welp there's
probably a decent chance the crew won't make it but we're going anyway
because if we don't <some easily enumerable bad thing will happen>", the
consequences to everyone involved were far too high to ever be actively
cavalier.
They had their processes and they followed the processes. Yeah Thiokol
engineers balked when asked about this particular launch but I expect
they balked relatively regularly it's no skin off their ass to say "no
go", but at the end of the day as a NASA-person your job is to fire
rockets with people on them from time to time, either have a manned
space program or don't. Can always find reasons not to launch.
The Thiokol engineers said not to launch below 56 degrees F, or the
SRB o-rings wouldn't seal. The temp was 19 that morning.
They were nervous, rightly so. But it’s not the outside air temperature
that matters, it’s the temperature of the o-rings.
The main issue is that what NASA had was a developmental system, with all
sorts of unknowns, and they were trying to run it as though it was a
well-understood production system.
Apollo had delivered amazingly—men on the moon in 9 years, starting with
modified ICBMs! The cultural impact of that was very large—not only did it
fix the US’s Sputnik problem and reassure their allies (which was the point
of the exercise), but it changed everyone’s attitude towards the Earth
itself.
To my eye, that “Earthrise” photo has had more lasting influence than all
the trudging around up there.
Naturally—NASA being a big government agency and therefore having
pathological incentives—they reacted by asking for the sky: enough dough
for a permanent space station, a Mars mission, and a fleet of space trucks
to get all that stuff to and from orbit.
Under severe budget constraints (for NASA, at least), they canceled all of
it except the truck fleet. Then they got some fancy consultants
(Mathematica Inc., no relation to Wolfram & Co.) to make a plan for 60
launches per year using 6 shuttles, with mostly commercial payloads, so
that they could afford to build them.
Eventually they managed nine launches per year, mostly military, and
instead of being everyone’s idol and independently wealthy, they were
begging for money and firmly under the thumb of the Air Force.
Meanwhile, they were trying to learn how to make and run the spacecraft.
They did it by the book, with a very highly organized system of flight
reviews and engineering criteria that allowed them to apply uniform
standards of evidence and decision-making to a wildly multifarious effort.
The workings of this system, and the people who made it work so well almost
all the time, are what Vaughan’s book is about, and what makes it so
fascinating.
What bit them in the end was that Thiokol wasn’t able to show a good
statistical correlation between O-ring erosion and the calculated
temperature. It got worse at low temperature, of course, but there was this
one launch where the rings were at 80 degrees or something, yet which had
the largest erosion to date.
Least squares analysis is very sensitive to outliers, so that one data
point destroyed the calculated correlation. The procedure required firm
data to delay the launch, and Thiokol didn’t have it. That’s how you run a
well-understood production system: You learn by doing, and make adjustments
based on accumulated experience. They were very good at that.
Because the shuttle was really a developmental system, though, the
procedures for making those adjustments actually pushed the SRBs closer to
disaster with each anomaly.
I can’t do justice to it here, but the way this whole team of highly
competent people, who followed all the rules of the system conscientiously,
nevertheless caused a disaster, is a story right out of Greek tragedy.
Cheers
Phil Hobbs
--
Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC /
Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics