Downloaded from https://www.everyspec.com t ,o where as %% vwmAa stnaddeviation 2A/,31emm Ab'deW, M geu (z . d ftestnadsnna % \ .m.* We riso introduce the ftolowing notation. \ I, - man, time to failure - as IA i.e . 1 , n~m. 4C % 7431 $w--Asdmvs---..T gauf (z) "Calf of the standard s-normal X(Gaussian) ditribution (ju#O, t. o1), the probability of failure) sow a eufc(z) Sf of the standard s-normal 02 distribution (the reliability; it is the complement of the gnuf(z). (the reliability) 3 redundun.y (1-out-of-2:G, hot standby); cach I 0 Cue . Two elements in active parallel At has an s-normal distribution of time to failure William H.
Von AlEan, Ed., Ri lit E,"',,inw, 0 194 with parameters po ., and j, ,a,. Define bv ARINC Resarch Corporation. Reprinted by permission - of Prentice-Hall. Inc.. Englewood Cliffs. N.J. Z, !-a, 2b ( -9) FIGURE 9-2. Survivor Functionv for Two P/rticular Systems With the Same MTF From Eq. 8-3, the probability of failure is 6 = guf (z ) gau(z) (910) Fig. 9-2. From the figure, the redundant To illustrate Case 6, assume '.hat the two system has the greater reliability up to Xt ft components, A and B, have the following p 1.75. After that, the improved single-element parmeters: system is the more reliable.
The point of intersection of the two functions will change ia = 300 hr ,, = 400 hr if more redundant elements are added, if the a = 40 hr 0b 60 hr (9-11) degree of element improvement varies, or if In order to evaluate the reliability of this standby redundancy is used. redundant unit at, say 350 hr, the following In redundancy applications, there is computation iE performeJ using Eq. 9-9: usually one time, say t, when the reliability of a nonredundant system with improved 350 hr -- 300 hr 1. 'I elements is equal to the reliability of a redundant system with less reliable elements. Z 350 ir - 400 hr = -0.833 (9-12) When t t, the redundant system has the = 60 hr greater reliability.
When > t, the Now refer to the tables of the s-normal improved-element system is superior. The distribution. choice of the system configuration depends Unreliability or probability of on the ratio of element life to mission time. failure =- gauf(1.25) gauf(-0.833) = 0.8944 X 0.2026 - 0.1812 -- 0.18 (9-13) 94 THE s-NORMAL D)ISTRIBUTION 9-5 OTHER CONFIGURATIONS The s-normal distribution is useful to describe many systems whose failure rate Table 9-2 lists the reliability of several increases "to infinity". Its pdf is combinations of elements. The last column shows the MTF under the assumption that all 1 e -1 t- elements have an identical constant failure f(t) exp - (--.- ] (9-8) rate. 9-3