Downloaded from https://www.everyspec.com AMCP 706-l0 Eampl o No. 1 (a) D~t a et A. Beta Bounds Method (Table 2--. Fig. 2-14B . 1. Plot the 5%/t0%/95% points for each failure 1. Prepare Table 2-3 col. 3 time. from Table 2-1(B). Plot on Fig. 2-1(B). 2. Sketch a curve through the 5% and through 2. See curves I and 5 on Fig, the 95% plotting points. 2-1(B). 3. Draw the "be't" straight line through the 3. Draw line 03. 50% plotting points. It will moreo'r4a.s bisect the region betmeen the 5% and 95%, plotting point curves. 4. Draw 2 more straight lines through a cen- 4. Draw lines 02 and 04. tral 50% population point:
one with the max- imum feasible slope and the other with the minimum feasible slope. 5. Estimi~te su, A, F 4 L from the curves 11, 5. U - 2460 hr (high limit) #3, 1b, respectively. They are all the in- 1960 hr (point estimate) tersection with the heavy dashed line. They are the high limit, point estimate, and low L - 1460 hr (low limit). limit, l espectively, 6. A separate scale is rarely given for the 6. oa - (3400 - 500)/2 m 1450 hr ,I scale parameter a. It is easily estimated 0 froM the fact that the 16% and 84% popula- a( - 1000) 990 hr tion pointh are each I a away from the 50% a - (2600 - 1300)/2 - 650 hr. population point.
The S-suale could have been used, alonj with the instructions on the book of Fig. 2-1(A). (au, &, and aL are read fronm ourves 2, 3, and 4 of Fig. 2-1(B), respectively.) It is not feasible to put quantitative "-oonfidenoe levels on the upper and lower limits for A or a. The conclusions to be drawn are essentially the same as fr'om Fig. 2-1(A). O4