GB/T 2829-2025Sampling procedures and tables for periodic inspection by attributes (English PDF)

3.1 Terms and definitions

A thing that can be described and considered independently. a system, or a combination thereof. A certain quantity of certain products, materials, or services that are brought together under certain conditions. production pattern inspection, or all items manufactured within a specified period during the mass production process after passing the production pattern inspection. The number of items contained in a lot. The item selected from the lot for inspection. A subset of a population consisting of one or more sample units. The number of sample units contained in the sample. units in the sample after the last stage of sampling is completed. The quality characteristics of the item do not meet the requirements. item quality represented by the quality characteristics, or according to the severity of the quality characteristic nonconformity – into: class A nonconformity, class B nonconformity and class C nonconformity. The critical quality characteristics of the item do not meet the requirements, or the quality characteristics of the item are extremely seriously non-compliant with the requirements. The major quality characteristics of the item do not meet the requirements, or the quality characteristics of the item are seriously non-compliant with the requirements. rejectable quality level (RQL) that is larger than that of a class A nonconformity but smaller than that of a class C nonconformity. The general quality characteristics of the item do not meet the requirements, or the quality characteristics of the item are slightly non-compliant with the requirements. quality level (RQL) than that for both Class A and Class B. An item with one or more nonconformities. be divided into: class A nonconforming item, class B nonconforming item, and class C nonconforming item. Class A nonconforming items are those with one or more class A nonconformities, or also with class B and/or class C nonconformities; class B nonconforming items are items with one or more class B nonconformities, or also with class C nonconformities, but without class A nonconformities; class C nonconforming items are items with one or more class C nonconformities, but without class A and class B nonconformities. The quality of an individually submitted lot (expressed as percent nonconforming or nonconformities per 100 items). The lower limit of lot quality considered unacceptable during periodic inspection. The activity of measuring, inspecting, testing one or more characteristics, and comparing it with specified requirements to determine whether a product or process is conforming. […]

3.2 Symbols

The following symbols apply to this document. number, change the numerical subscript c to 1, 2, .... For example, A2 represents the second acceptance number. sample size, etc., add numerical subscripts 1, 2, ... respectively. For example, n5 represents the fifth sample size. […]

5.4 Classification of nonconformities

Based on the specific circumstances of each test group, the nonconformity of quality characteristics shall be classified into three classes: A, B, and C. […]

5.5 Specifications for rejectable quality level

RQL is a quality indicator used in periodic inspection by attributes to assess the stability of a production process. When the rate of nonconforming items in the production process is greater than the RQL value, the sampling plan shall judge with a high probability that the production process is unstable (periodic inspection unacceptable); when the rate of nonconforming items in the production process is much less than the RQL value, the sampling plan shall judge with a high probability that the production process is stable (periodic inspection acceptable).

The rejectable quality level shall be determined by the user and the manufacturer through consultation in the product technical standards or purchase contract. In principle, the rejectable quality level shall be specified separately for each test group. Alternatively, it can be considered to specify rejectable quality levels for different classes of nonconformities within the same test group, or to specify rejectable quality levels between different test groups or between the same class of nonconformities in different test groups.

Generally, the rejectable quality level specified for class A shall be smaller than that specified for class B, and the rejectable quality level specified for class C shall be larger than that specified for class B. When the rejectable quality level is expressed as a percent nonconforming, the RQL value shall not exceed 100. When the rejectable quality level is expressed as the nonconformities per 100 items, the RQL value shall not exceed 1 000. For a method to determine the RQL value, refer to Appendix A.

In Tables 2~22, the rejectable quality level less than 100 can be either a percent nonconforming or the nonconformities per 100 items; the rejectable quality level greater than or equal to 100 shall be the nonconformities per 100 items only. In addition, the rejectable quality levels given in these tables are priority values. If the specified rejectable quality level is not a priority value, these tables are not applicable.

5.6 Specifications for determinant level

This document presents three determinant levels of different capabilities. The higher the determinant level, the higher the requirement for the stability of the production process and the lower the probability of passing the periodic inspection. Among them, the determinant level III is of the strongest capability, followed by the determinant level II, and then the determinant level I. When strong determinant capability is required and it is economically feasible, adopt determinant level III; […]

5.8 Search of sampling plan

5.12 Procedures after periodic inspection

6.1 Operating characteristic curve (OC curve)

Figures 2~21 in this document, ordered by sample size 1 for single sampling plan with the same sample size, present the operating characteristic curves of single sampling plans (including three determinant levels) based on binomial distribution (suitable for inspection of percent nonconforming, rejectable quality level less than 100%) and Poisson distribution (suitable for inspection of nonconformities per 100 items, rejectable quality level less than 1 000). The x-axis of the operating characteristic curves given in this document is scaled by the lot quality p, and they correspond one-to-one with the operating characteristic curves of the single sampling plans in Tables 2~4. The arrays in square brackets represent single sampling plans for inspection of percent nonconforming, and the arrays in parentheses represent single sampling plans for inspection of nonconformities per 100 items. Meanwhile, the operating characteristic curves of the double and quintic sampling plans in Tables 5~10 are essentially consistent with the operational characteristic curves of the corresponding single sampling plans.

This document presents the values of Pa (RQL) for single sampling plans of three determinant levels based on the binomial distribution (corresponding to the table value for inspection of percent nonconforming) and the Poisson distribution (corresponding to the table value for inspection of nonconformities per 100 items). It also provides the values of corresponding lot quality p for the single sampling plans with probabilities of acceptance of 99.0%, 95.0%, 90.0%, 75.0%, 50.0%, 25.0%, 10.0%, 5.0%, and 1.0%.

6.2 Average sample size curve (ASN curve)

Figures 22~38 in this document, ordered by the decision array [Ac, Re] of the single sampling plan, present the average sample size curves of the double and quintic sampling plans of the design value corresponding to the single sampling plan of the design value on the same figure based on the Poisson distribution. The x-axis of the average sample size curves given in this document is scaled by the lot quality p (the number of nonconforming items out of 100 items). The average sample size curves of the double and quintic sampling plans in Tables 5~10 corresponding to a certain single sampling plan in Tables 2~4 of this document, obtained according to the method given in Appendix B, are almost identical to the average sample size curves of the double and quintic sampling plans based on the Poisson distribution in Tables 5~10. […]