Downloaded from https://www.everyspec.com AMCP 70-6S$ Pr probability of event a - producer risk in . 0 = consumer risk r = number of failures 7' - discritnination ratio/ ! r* number of failures for ,dsmni ro rejection 0 I I/A, exponential scale parameter. mean of R =s-reliability, Sf failure time . = s-unreliability, R I --. R, wrconstant failulvte, a Cdf Poisson rate' parametei RL - reliability level JA = S, St slope associated with " value of chi-square such accept and reject decision that csqf (4•,; P) - P lines; see Eqs, 3-12 and 3-28; number of test 1v0" = subscripts:
0 implies stations ,."" H 0 , 1 implies H, 3- Implies the word I-sample a single-sample plan "statistical(ly)", or./ "implies that the statisti- 3.1 INTRODUCTION cal definition is intended rather than the ordinary After an equipmeni or system design is dictionary definition well established, reliability-testing changes from a design tool to a tool for making seq-sample a a sequential-sample plan decisions and for determining if reliability goals have been met. These tests range from Sf - Survivor function, quality assurance tests, which are performed Sf 1 '- I - Cdf{.
at the parts level on submitted lots, to rell- for a continuous variable ability demonstration tests at the system level (Ref. 1). T = total test time A test program that produces spurious T* = value of T for acceptance results can lead to accepting unreliable equip- ment. The cost of spare parts and extra TM m mission time, see par. maintenance can far exceed the cost of 3-6.3 accurate testing. Therefore, a carefully de- signed test program which considers s-con- URL - unacceptable reliability fidence levels, sample sizes, consumer and level producer risks, and test cost must be devel- oped.
In any reliability test, many engineer- Wt waiting time before deci- ing factors must be considered in addition slon, see par. 3-6.3 to statistical factors. Z a standard s-normal Often, it is difficult to perform a reliabil- variate, par. 3-8 ity demonstration test on a complete system, ..... . . -.. I -- ~ ~ 2----.----,.,,-.