Originally Posted by
mike-wsm
[...] Many thousands of highly educated aerospace engineers have put immense amounts of effort into making the aircraft multiply redundant and failure-tolerant so that any one failure or set of failures can be survived.
Airframes have single points of (catastrophic) failure, just like engines.
How many HS's do you have, and how well can you fly if one falls off ? How many main wing spars ? Your cargo door seals itself when the fuselage is pressurized - or is it just held closed by a latch (I mean, why on earth would an airframe engineer do that and not use a plug-door?), and what if that latch, or its controller, fails ?
Or even just the control surfaces. Say, uncommanded rudder hard-over. Your airframe designers would have ensured no single point of failure could cause that, or that the other control surfaces were sufficient to allow the pilot to handle it, right ? Or if, lets say, they used a jackscrew assembly that
was a single point of failure, then they'd make sure that jackscrew was itself failsafe with redundancy (like NASA did), right ?
Reality is that no airframe or engine that is completely redundant and failure tolerant would ever get off the ground (literally). Safe, but useless. Engineering compromises
have to be made, and sooner or later one of those
will kill. The aim is to make it later rather than sooner, but engineers, like pilots, are human and don't always get it right.
Final note: IMO in any branch of engineering, any belief that you have
elimiinated the possiblity of failure ["
any one failure or set of failures can be survived"] is asking for a swift kicking from Mr Murphy & Laws...