Originally Posted by
SyEng
Avrflr,
“Miracle” is not standard systems safety analysis terminology. Please define. Here are some of the terms we use in the industry by way of example. Personally though, I prefer to stick to numbers for clarity.
Remote: < 1E-5 failure events per flying hour
Extremely remote < 1E-7 “
Extremely improbable < 1E-9 “
A miracle: the odds of it happening within the lifetime of the Universe is indistinguishable from zero. Probability that I may be proven to be totally wrong: 1 in 4
I accept that the accident was unlikely and therefore had an unlikely cause. I don't believe that gives one carte blanche to suggest miraculously improbable coincidences as the cause of the failure.
Originally Posted by
SyEng
Can anyone out there help with an applicable list of FDR parameters?
That might be a useful exercise. We can agree that something did happen, and it appears to have gone unnoticed by the FDR. Perhaps a gap can be found where no data is collected that could provide a place for the problem to "hide".
The trouble with playing this game is that we are working with only partial information. If we knew exactly what the AAIB had tested and how they had tested it, we could come up with better theories, and for that matter, better arguments against theories. I may very well be making false assumptions as to what has been established as fact, based only on the very brief reports I have read. I don't believe any of the theories I have read are credible, based on the evidence in front of me. It seems that this is the conclusion the AAAIB have come to (with much more evidence), hence the continuing investigation.