Re: What was the biggest problem for each of the 2 destroyed US space shuttles?

From: rk (stellare_at_nospamplease.comcast.net)
Date: 09/14/04 

Date: Tue, 14 Sep 2004 02:10:26 -0500

Jay Windley wrote:

>
> "rk" <stellare@nospamplease.comcast.net> wrote in message
> news:Xns9562EC38A2C79rk@216.196.97.136...
>|
>| Here's what Boisjoly said in '87, in part. Others have similar stories
>| of being ignored.
>
> And those others are those same engineers who so confidently, and with
> great apparent technical justification, had laid out a case for nearly a
> decade that the joint was safe to fly even though it was not performing
> as designed.
>
>| Why try to change the topic? We are talking about a specific event.
>
> *You* are talking about a specific event. I am trying to broaden the
> field of view and interpret that event in terms of what happened in the
> years that preceded and followed that event. I believe that context is
> important to understanding why people did what they did that night. Not
> knowing why people acted the way they did runs the risk of repeating the
> same mistakes, only with different people.

They were ignored, you commented on that, and I commented on your comment.
Now you are trying to broaden things.

> As the Columbia accident demonstrated, the problem is not about
> individuals or individual wrongdoing, but about a culture that prevents
> people from doing the best job they can.

It's about both. Placing the full blame on culture is ridiculous. The
culture can make the decision to speak up in a meeting more difficult but it
doesn't change any one's actions; people make decisions about their own
actions and control their own destiny in chosing whether to speak or to stay
silent. It's about both culture and character.

> This same culture was a problem
> for Challenger. Pointing to Boisjoly's disenfranchisement on the eve of
> the launch and the alleged management wrongdoing exposes the symptoms of
> the problem but does not address the underlying cause.

Boisjoly, Thompson, and McDonald were all ignored that night which is a
fundamental management error.

>| If you want to talk about other launches fine but you appear to be
>| running away from your position. Fools no one.
>
> I'm not trying to fool anyone. And I will gladly abandon my position if
> it is shown to be materially in error. I'm not trying to be right for my
> own gratification. I'm trying to be right because I believe us being
> (collectively) right helps us have a safer and more productive space
> program.

Your position is all over the place (see your discussion on the normalization
of deviance of a bad motor in a car, where you take both sides at once and
surround it with lots of words taken from some books).

And you are ignoring a lot of the data that is out there. I have added some
for you to consider.

 
>| >| All your words below (and I don't have time right now to go
>| >| through all of that as there are a lot of errors) ignores the
>| >| fact that the onus is to prove safety, not to prove there is a
>| >| problem...
>| >
>| > I covered that in a previous post.
>|
>| OK, looks like I'm done responding to you unless you have something
>| meaningful to say.
>
> To summarize the post to which I alluded, the culture was less about
> arguing against the default not to fly and more about requiring
> contractors to provide full technical rationales for their
> recommendations. That's what the whole NASA quality system was designed
> to achieve.

I'd like to see your citation for that. I was under the impression that the
whole NASA quality system was not designed to provide full technical
rationales for their recommendations but that it was designed to provide for
full technical rationales supporting an OK to launch. You are making the same
mistake that the managers made in '86, blurring things in the gray zone. The
gray zone, from an engineering point of view means do not fly. An MRB may
decide otherwise and engineers will support that with the best technical
analysis that they can.

> The presidential commission -- and subsequently the public
> -- misunderstood terms like "launch constraint" and "waiver" that had
> precise meanings to NASA and its contractors. This has given rise to
> mistaken interpretations of the culture that prevailed at the time.

I think I know the meaning of constraints and waivers. I could be wrong so
I'll keep on reading.

> It was assumed that the contractors would recommend to fly, but have
> secret reservations. Thiokol had for many years provided a detailed
> rationale for flying with the SRB joints behaving as they did. This
> rationale was backed up by flight experience, by Thiokol's tests, by
> MSFC's tests, by a mathematical model, and by reference to the original
> design rationale (i.e., safety in redundancy). That kind of engineering
> rationale was not only expected, but also demanded, by those at NASA who
> supervised Thiokol.
>
> Boisjoly: "It was our feeling at the time that nothing gets presented to
> [MSFC director] Dr. [William] Lucas unless the people that are doing the
> presenting are absolutely sure that all bases have been covered
> technically and that they have all the answers to all the questions that
> potentially could be asked, and this was a case [i.e., temperature effect
> on SRB field joints] where clearly there were no answers available
> because it was just a question of observation as to what we were
> presenting." (Diane Vaughan, _The Challenger Launch Decision_, p. 222)

And that is not how the NASA system is designed to operate, from what I
understand.

 
> The culture of rigor, like the culture of learning by doing, also works
> both ways. If you rigorously support one point of view, and that rigor
> has come to be demanded, then changing the point of view requires equal
> -- nay, greater -- rigor.

No.

Again you are repeating the mistakes made in '86. That's not how engineering
works for any mission or safety critical systems.

I think you spent too much time reading Vaughan who can't even get the units
straight for MCO *after* the failure.

> You must not only show why your new viewpoint
> is technically justified, but also what is wrong with the previous
> rationale.

Again you are making the same mistake as was made in '86.

> It has to be kept clearly in mind that Thiokol's attempt to
> provide data to support their recommendation not to launch, backfired.
> The data they presented to NASA on the eve of the launch supported a case
> *against* a supposed adverse effect of temperature on O-ring performance.

Nope. It supported a case not to launch. Vaughan didn't understand
engineering at all and was incapable of performing any analysis on what she
read. The data could have been presented better, and many agree on that.
Forcing real-time presentations and reviews was an additional serious
management mistake. While that satisfied people like Vaughan so that they can
claim that "procedures were followed" and "no rules were broken" that is
simply saying that format matters and content does not. If you really want to
understand Vaugahn, you may wish to note who actively supported the writing of
that book and who was mostly ignored.

> All they had was the general argument -- true though it may be -- that
> elastomers lose their elasticity in the cold.

No, they also had data that they had taken and their calculations.

And they also had on their side that they could not back a launch decision.
It was in the gray area. Ignoring this was a serious managemenet mistake.

And they also had the pictures of damaged joints they pounded on the table.

> But they had no data that
> the elastomeric O-rings in the SRB joints (i.e., that a specific
> elastomer in a specific context) followed that general rule, and indeed
> evidence to suggest that it did not.

Really. 

-- 
rk, Just an OldEngineer
"Engineers abhor extrapolation"
-- Ken Iliff, from _Runway to Orbit_, 2004

Relevant Pages

  • Re: What was the biggest problem for each of the 2 destroyed US space shuttles?
    ... Placing the full blame on culture is ridiculous. ... > rationale for flying with the SRB joints behaving as they did. ... If you rigorously support one point of view, ... Again you are making the same mistake as was made in '86. ...
    (sci.space.shuttle)
  • Re: Wood is expensive
    ... started absent evidence of the stated rationale have served to ... strategic purposes we deny despite the decades of evidence to the ... Petroleum and US Petroleum corporations - offering moot testimony as ... Bush has the secret support of the terrorists whose ...
    (rec.woodworking)
  • Re: What was the biggest problem for each of the 2 destroyed US space shuttles?
    ... but about a culture that prevents people from ... the whole NASA quality system was designed to achieve. ... "launch constraint" and "waiver" that had precise meanings to NASA and its ... This rationale was backed ...
    (sci.space.history)
  • Re: What was the biggest problem for each of the 2 destroyed US space shuttles?
    ... but about a culture that prevents people from ... the whole NASA quality system was designed to achieve. ... "launch constraint" and "waiver" that had precise meanings to NASA and its ... This rationale was backed ...
    (sci.space.shuttle)
  • Re: Problem with Decimal.Parse and CultureInfo
    ... I've already tried to set UseUserOverride to false, ... if french culture were not present in the framework. ... Microsoft MSDN Online Support Lead ... where an initial response from the community or a Microsoft Support ...
    (microsoft.public.dotnet.framework.aspnet)