A few important thoughts on failure modes and distributions, recently discussed on Linkedin platform.
Components fail either because they do wear out (right side of the bathtub curve), and it is possible to estimate their safe life limit or economic life limit (ref. Moubray RCM II). Or they fail because quality defects in the manufacturing or installation process cause them to fail during burn-in phase (left side of the bathtub curve).
Or they fail during lifetime once when the actual applied stress exceeds the inherent strength of the component (so called “random failures” or the middle section of the bathtub curve). It really is as simple as that. But that is just on the theoretical side of the discussion.
When we come to the reality and the operating context, this is what often makes the whole difference: random failures can turn into wear-out, and wear-out can turn into random, depending on how the equipment is operated. Remember this statement because it emphasizes the No.1 reason for the failures (or lack of effectiveness) of RCM analysis: an incomplete awareness and consideration of the operating context. Make sure it is fully stated, approved by all parties, and documented.
