SRC Forum  Message Replies
Forum: Reliability & Maintainability Questions and AnswersTopic: Reliability & Maintainability Questions and Answers
Topic Posted by: Reliability & Maintainability Forum
(src_forum@alionscience.com
)
Organization: System Reliability Center
Date Posted: Mon Aug 31 12:47:36 US/Eastern 1998
Original Message:
Posted by: Clay Davis
Date posted: Wed May 24 0:35:57 US/Eastern 2000
Subject: L1 to MTBF Conversion
Message: Is there a conversion factor between a L1 number and a MTBF number? Or a generic L# and MTBF? L1 is when 1% of the population will fail. Also, is it correct to say MTBF is equivalent to L50?
Thanks. Clay
Reply:
Subject: Li to MTBF Conversion
Reply Posted by: Patrick T. Hetherington
Date Posted: Wed May 24 6:28:43 US/Eastern 2000
Message: There is not a generic conversion. The figure L1 or more generally Ln is a time to failure distribtuion independent term representing the % of a population that came be expected to failure by a given time. Ln alone gives no insight as to when the n% will fail or how long the 100n% will last, simply that n% will have failed by a given Ln time. Ln is a term generally used when speaking of nonrepairables items and gives a point estimate of reliability. Non repairable items may or may not have time to failure distributions that are random(constant # of failures/unit time). MTBF is a term generally reserved for repairables systems that have times between failures randomly distibuted. In the general sense, there is no relation to Ln for a item and the MTBF for a repairable system.
For nonrepairables items, the time to failure distribution can most often be approximately by the Weibull distribtuion. Other possibities include normal and lognormal distribtions. For the Weibull distribution the relationship of MTTF=thete*gamma funciton(1+1/beta)+To where beta and theta are paramters of the distribution determined by curve fitting or with software. To is interval beginning at time =0 before which no failures are expected (often = 0). MTTF represents the average of all expected times to failure of a distribtuion. This can be wrongly interpreted if that distribution of failures is not symetrical. For nonrepairable units that are normally distributed (symetrical about mean), the L50 life is equal to the mean and the MTTF. For nonrepairable items that have a random failure distibution, the MTTF corresponds to the L63 life ( a lot fail before the average time to failure and some last a really long time). If the opposite were the L40 life may be equal to the MTTF.
