||Taylor Enterprises, Inc.
|Quality and Statistics
Books, Software, Training and Consulting
Sampling Plan Analyzer Help System
Contents, Index, Glossary,
ANSI Z1.9 for Defective Units
The ANSI Z1.9 standard can be used to inspect for defective units when dealing with a measurable characteristic that is compared to specification limits. ANSI Z1.9 contains tables of variables single sampling plans for defective units. It was based on MIL-STD-414 which has been obsoleted.
AQL = 0.010%, ..., 10%.
Level Of Inspection = S3, S4, L-I, L-II and L-III
Lot Size = from 2 on up
Type = Standard Deviation Known - 1 spec - k (Form 1)
Standard Deviation Known - 1 spec - M (Form 2)
Standard Deviation Known - 2 specs - M (Form 2)
Standard Deviation Unknown - 1 spec - k (Form 1)
Standard Deviation Unknown - 1 spec - M (Form 2)
Standard Deviation Unknown - 2 specs - M (Form 2)
Options = normal, tightened, reduced, full switching, switching without reduced inspection
For a given AQL, level of inspection, lot size and type, the standard gives three plans called normal, tightened and reduced. You can select an individual sampling plan from the standard and evaluate it. The standard also contains a set of rules for switching between these plans. The standard intends that these switching rules be used. You can also evaluate the combined protection obtained using these switching rules. The switching rules work as follows:
Start: Start using the normal sampling plan.
N to T: If, while on normal inspection, 2 of 5 consecutive lots are rejected, switch to the tightened plan.
N to R: If, while on normal inspection, 10 consecutive lots are accepted, switch to the reduced plan.
T to N: If, while on tightened inspection, 5 consecutive lots are accepted, switch to the normal plan.
R to N: If, while on reduced inspection, a lot is rejected, switch to the normal plan.
The use of the reduced sampling plan is optional. The exact rules are displayed in the ANSI Z1.9 for Defective Units dialog box once you have selected the plan.
(1) Representative sample
(2) The measurements are normally distributed.
(3) For the standard deviation known plans, it is assumed the standard deviation is consistent over time.
(4) If the switching rules are used, it is assumed that one is inspecting a series lots.