Implications of MTBF

[J. Bridges]

Does using the term "MTBF" imply that the population one is referring to has a constant failure rate?

[F. Valera]

Ideally speaking, MTBF is used only to repairables item but normally everybody used it for both repairable and non-repairable item. On the other hands, you are right but not entirely correct because that suppose population with following a exponential distribution (constant failure rate). We shouldn´t forgett that MTBF is an statistical value, is the MEAN TIME (or the average) between failures.

[RS Support]

It can, but it depends on what you mean by MTBF. If you are referring to Mean Time Between Failures, then yes. However, if you are referring to Mean Time Before Failure (or MTTF) then this does not imply a constant failure rate. The only time that the Mean Time Between Failure translation is correct is when the underlying distribution has a constant failure rate (e.g. exponential). If the failure rate is constant then Mean Time Between Failure and Mean Time Before Failure are indentical. However, if the failure rate is not constant then these two terms are not equal.

[Annandhan Mayappan]

MTBF - If it is Mean Time Before Failure, then is is not a constant

MTBF - If is is Mean Time Between Failure, then it should be contant in general