Predicting Mechanical Component Reliability

Ah, Weibull Analysis... takes me back to my Mechanical Reliability Course taught by Professor Rivera at RPI (Rensselaer Polytechnic Institute, Troy, NY.) Good times.

Do the airlines use any of these methods to predict part failures, or do they use a different technique?

Yes, and/or yes. Which is, in part, to say: these are not, per se, so much techniques as they are, simply, standards established for the design and manufacturing process...a systemetized way to go about assuring the processes of reliability assurance at the various life cycle stages of a product. There is nothing "sacrosanct" in the techniques themselves. Seen from another angle, the "methods" do not, themselves, "determine" part failure rates--that is done by various kinds of testing, both destructive and non-descructive. It is the results of such tests, as compiled in databases, that are incorporated into the various system design validations in order to "predict" a likelihood of non-failure (of "failure success") of a design and product throughout product life cycle. Also standardized is the means of reporting the processes undertaken to fulfill customer requirements as they relate to system reliability. Again, the importance is in standardization...so the reporting process does not become haphazard; or have to be redefined, in detail, with each RFP for each proposed project.