View Single Post
  #5  
Old 06-05-2009, 10:45 PM
Star Trek Viewer Star Trek Viewer is offline
Banned
 
Join Date: May 2009
Posts: 1,057
Default

Quote:
Originally Posted by Akula2ssn View Post
Personally I prefer fail-safe, but safe-life can have applications.

In a safe-life design, the product is designed to survive a specific "design life" usually with a reserve built in. So typically Airbus planes are designed to function for approximately 20 years. After which the plane is to be replaced, however the drawback is that the plane could technically probably maintain service longer. The idea in a safe-life design is that a critical system is designed to work for years without the need for repairs, this is usually done in systems that are difficult to access and/or repair.

Fail-safe is a design philosophy where should a failure actually take place in a component, the system as a whole can respond in a way to minimize harm to other components or personnel. So basically if a component were to fail, that one failure should not lead to the destruction of the whole system. This is also similar to "fault-tolerant" designs.

This is why generally you don't normally see Airbus planes in service that are over 20 years old. There are probably exceptions.

In a way, I guess you could say that in safe-life, the component isn't supposed to fail in the first place. Fail-safe takes steps to be ready in case a failure does take place.

But like I said earlier, you can design the hell out of something but there are still limits to what a design can take, even one that uses a fail-safe design philosophy.
Thanks for that explanation.

I think that fail-safe philosophies better take into account that one cannot really know the actual life of an individual component. For example, a widget can have a design life of 20 years, but what does that really mean? What if the widget in question has hairline cracks in it that x-ray inspections cannot catch? Then, in such a case of defect, the actual life of full reliability might only be, say, 10 years, and the last 10 years are basically unsafe-life in practice.

By comparison, in case of either a design or an individual defect, fail-safe philosophies allow for unanticipated failures such that in case of such failures, the risk to the larger system is minimized.

However, I have no independent knowledge of what Airbus's design philosophy really is, and I certainly am no expert in this area.
Reply With Quote