Downloaded from https://www.everyspec.com other, as the number of defects listed under one result in hazardous or unsafe conditions for iA,- classlication increases, the probability that all divdials using or maintaining the prcduct; or, of them will occur with the same frequency de- for major end-item untits 4 i product, such as creases. ships, aircraft, or tanks, a defect that could prevent performance of their tactical function. When setting an AQL for a component part it is c. Major Defects.
A major defect is a de- important to view the item as a whole, in order fect, other than critical, that could result in that the importance of different features can be failure, or materially reduce the usability of seen in proper perspective. For example, con- the unit of product for its intended purpose. sider the case of a squib that is to be a com- d. Minor Defects. A minor defect isonethat ponent of a rocket. The AQL for the functioning does not materially reduce the usability of the test of the rocket is determined to be 2.5 per- unit of product for its intended purpose, or is a cent.
This means that the AQL must be more departure from established standard3 having no stringent for the functioning test of the igniter significant bearing on the effective use or op- (1.5 percent) and still more stringent for the eration of the unit. functioning test of the squib (0.65 percent). This must be done in order to avoid penalizing the 5-9. Inspection by Attributes is the grading of a manufacturer of the igniter (the functioning of unit of product as defective or nondefective, with which depends on the squib) and the manufac- respect to a given requirement, by sampling in- turer of the rocket (the functioning of which de- spection.
When attributes inspection is per- pends on the squib and the igniter). However, formed, the decision to accept or reject is based if it is known that the same squib is to be used on the number of sample units found defective. in a JATO unit, for which the functioning test The extent of deviation from the requirement is has an AQL of 0.25 percent, the AQL for the not considered in this form of inspection. This ,' squib in this case should be about 0.10 percent.
form is simple to administrate, normally is Obviously, the same squib cannot have two simple and rapid to perform, is not dependent AQL's; therefore, the squib intended for the upon a particular distribution of product to pro- • r.rocket, also must have an AQL of 0.10 percent, vide, a desired• assuratnce, and inrolves no math - unless a method of grading squibs can be de- ematics to determine lot acceptability. It is al- vised. In all cases the component snould be ways possible to use attributes inspection, and given the most stringent AQL required for any in many instances, such as in GO and NOT GO use. gaging, it is the only type possible. 5-8. Classification of Defects.
Defprtsofsimi- 5-10. Single-Sampling is a technique in whi.h lar importance are treated alike. To accom- only one sample of n items is inspected to reach plish this it is iecessary to classify them into a decision on the disposition of the lot. A sam- groups, the number of which is a compromise pling plan is best described by the sample size between the degree of selectivity desired and (n) and ,cceptance number (c). When the num- the administrative complexity.
The following ber of defectives found in this sample equals, excerpts taken from MIL-STD-105 reflect the or is less than, the acceptance number pre- prevalent classes and definitions for the scribed by the sampling plan, the lot is accepted. classes. If the number of defectives exceeds the accept- a. Method of Classifying Defects. A classi- ance number, the lot is rejected. An example of fication of defects is the enumeration of pos- a single plan is as follows. Select a sample of sible defects of the u.,iit of product classifie' 100 units. If 3 or fewer defectives are found, according to their importance.
A defect is ax accept the iot; if more than 3 are found, re- deviation of the unit of product from require- ject the lot. ments of the specifications, drawings, purchase descriptions, and any changes thereto in the Single plans may be summarized as follows: contract or order. Defects are normally grouped into one or more of the following Sample size Acceptance no. Rejection no. classes; however, the Government reserves 100 3 4 the right to group defects into other classes. b. Critical Defects. A critical defect is one 5-11. Double-Sampling Is a technique in which that judgment and experience indicate could a second sample of n items is inmpected when 5.s