00315 Ã2897da0df33ebfef7d49591ddf4696

318

McCarviIle & Montgomery

FigurÄ™ 3. The Probabiiity of Having Some Defective Units in tbe RangÄ™ dx and tbe Associated Probabiiity of Those Reject Units Being Accepted by

tbe Gage.

P(reject)good) = J g(x)dx    (6)

USL

FigurÄ™ 4 shows this relationship. Again, an important eÄ…uality for this new risk is

P(reject|good) = jg(x)dx = jg'(x)dx    (7)

USL    -«

where g(x) ~ N(xj,crg) and g'(x) ~ N(USL,crg). This relationship is easily seen in FigurÄ™ 5.

In a similar fashion to that for calculating the defect levels, the probabiiity of having some good units as well as the probabiiity that these