Downloaded from https://www.everyspec.com AMCP 70&196 were line #31, there could be an appreciable where number of early failures, although the picture doesn't look as bad as It did on a-normal a -soule parameter paper (Fia. 2-1(A)). P *shape parameter Data Sets B and C, from Tables 2-4 and 2-5 ane plotted in Fig. 2-5; the r~ 1. T Rounds The distribution 13 discussed more fully in method is used, with 90% s-corat-enoe par, 2-3.6 and In Part Six. Mathematical buunds (N - 10, KS,, a 0.37). Data Set B Appendix and Gloaaarj', is the lower half of the plotted points (circles); Date Set C is the entire set of Figs.
2-7, 2-8, and 2-9 show Data Sets A, points, Fobr Data Set B, with only half the [text continues on page 2-361 po~ints, the uncertainy in the distiutiuon appears tremendous. The lines ane drawn for TABLE 24 Data Set C, the full sample. Following the instructions on the reverse uide of Fig. 2-5, DATA SET D one obtains This Isea simulated data sut from random-number I ck 700 hr, 1150 hr, au -1500 hr tals N 21. 0L . 5 5 , u2. 0, PU - 3 . 6 ~~n~eUn, Tdik 2.110) be. Tids snample could easily have comne from a Time to twom N -20 Weibull distribution.
The characteristic life Has* faillre, hr r/N seid N a 22) (at)Is known to within afactor of 2 -thde B 0 - f- life (Cdf a 10%) is probably between 10 hr I 816' 4.8 11,25M.3./13 and S(00 hr. 2 00.8 9.5 1.7/7.9/21 3 107 14,3 4.1/12/26 Data Set D (Table 2-6) was simulated from 4 135 2,6 6.10/22/34 a table of pseudo-randomn numbers. The 6 141 238. 14/27W procedure for plotting it on Weibidl probabil. 7 152 286 14/271/4 ity paper using the K-S Bounds method is 8 16 X61 20/"66 shown in some detail; see Example No. 1(F).
9 162 42.9 22/41/00 10 181 47,6 27/5/6 2-2.5 LOGNORMAL D1STRIBUTION 11 206 52.4 311/0/1111 12 206 57,1 30/111/73 Thiere are several forms for writing the 13 234 61l9 41/60/77 lognormal distribution-some of thern are 14 240 66. 46/64/90 confusing because of a carryover from the 16 245 71.4 51/00/a6 i-normal distribution of the (PA~a notation. 17 247 76. 613/76 One never quite knows whether IA and a 18 261 65.7 611111093111 refer to mean and standard deviation at all, 19 279 905. 72/67/6.9 mud if they do, whether it is to the lop or 20 279 95.2 79/92JAIII.3 not.
'The form used here is 21 281 100.0 07/96.7/111111.715 lognormal Cdf it, a, PI} Column 4 Is shown only for completeness; It Is not Illustrated with a figure. -gauf (in ((/cr)] (2-1~3) 1 - I IN 0.962, use 95%, s-wnf ldmnc. KS,, 11029 .........