creating a components reliability for a aircraft operator

[pannam]

Hi,
we are a small fleet aircraft operator. We have been operating since the past 16 years. Recently we were asked to prepare a components reliability.

Components reliability is faily new to us and frankly speaking I dont know where to start.
I am able to gather enough information of components and calculate MTBF, MTBR, MTTF, MTTR , FAILURE RATE AND UCRR. I have enough data but i dont know what to compare this with ? How do i establish an alert level?

 

[Dougt115]

What kind of alert level are you trying to create?

Reliability requirements are predicated on a desired probability of surviving a mission. Be it operating hours, cycles or just time.

The reliability requirement usually provided and the reliability analysis demonstrates through analysis or test that the unit as designed will meet the goal. An alert level implies that you want to warn a user when a certain criteria has been reached be it cycles completed, hours of operation, etc.

Please provide more details.

 

[pannam]

With 2 years of data I have calculated the MTBUR and other Means as mentioned in the first post. Now, for the unscheduled component that are removed every month I will need to compare it with something. i am confused on this part.
lets say a component "generator" is removed from an aircraft having done 3000 hrs. How do i know if the removal hours is comparitively good or bad. ? With the data's that i have wouldn't it be possible to set up an alert level such that when a component is removed (on a monthly basis) i can then plot a graph for reference or easy analysis ?

 

[IRstuff]

We must be having some sort of language problem. If a part is not failed, there's not much to compare against except total hours or hours since last repair. Nevertheless, in your example, if a monthly pull shows 3000 hrs since the last month, there's something else going on, given that there are only ~744 hours in a month.

Note that MTBF implies a reliability of 37% when that time has elapsed. Also, for an aircraft, there are fatigue and wear mechanisms that are not properly addressed by MTBF, per se.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[Dougt115]

Using the MTBF and an acceptable confidence level you should be able to determine the appropriate time to remove a unit from service.

Confidence level being the probability that your answer is within a specified range. The lower the confidence number, 60%, 50%, 20% the less likely your number is within the range. Ideally you would like a narrow range and a large confidence number. This requires a lot of data and a mature design.

Knowing the MTBF, operating hours, and the requirement for Probability of success, the useful life before failure of the unit can be determine. This is often determined in operating hours and duty cycles (percent of time the unit is actually operating) and an error/margin. Then just track the hours on a graph or spreadsheet to determine at which service interval to replace the unit. You should be proactive in this as the unit's useful life may end before the next service interval. Meaning the unit may have to be replaced short of the predicted useful life.

(I am assuming Aircraft cannot wait for failures to occur in actual applications as this could prove fatal.)