Regulatory Guide 6.6
| ML003739246 | |
| Person / Time | |
|---|---|
| Issue date: | 06/30/1974 |
| From: | Office of Nuclear Regulatory Research |
| To: | |
| References | |
| RG-6.6 | |
| Download: ML003739246 (7) | |
{{#Wiki_filter:U.S. ATOMIC ENERGY COMMISSION REGULATORYJune 1974GUIDEDIRECTORATE OF REGULATORY STANDARDS REGULATORY GUIDE 6.6 ACCEPTANCE SAMPLING PROCEDURES FOR EXEMPTED AND GENERALLY LICENSED ITEMS CONTAINING BYPRODUCT MATERIAL
A. INTRODUCTION
Part 32, "Specific Licenses to Manufacture, Distribute, or Import Exempted and Generally Licensed Items Containing Byproduct Material," of Title 10 of the Code of Federal Regulations requires certain minimum quality assurance practices for exempted and generally licensed items containing byproduct material, including the use of acceptance sampling. Section 32.110, "Acceptance Sampling Procedures Under Certain Specific Licenses," specifies acceptance sampling procedufes for use under certain specific licenses. This guide describes certain information needed by the Regulatory staff in its review of applications for licenses and provides guidance concerning alternative sampling plans that are acceptable to the Regulatory staff.
B. DISCUSSION
The purpose of the acceptance sampling procedures specified in §32.110 is to limit the risk that inspection lots of devices of excessively poor conformance to specifications will reach the public. Of the various criteria for selecting plans for acceptance sampling by attributes, lot tolerance percent defective (LTPD) is the most appropriate for this purpose. The following definitions* are relevant to this guide: *ASQC STANDARD A2-1962, "Definitions and Symbols for Acceptance Sampling by Attributes," American Society for Quality Control. Copies may be obtained from the American Society for Quality Control, 161 W. Wisconsin Avenue, Milwaukee, Wisconsin 53203. This document is the source for standard definitions used in this guide.1. Lot Tolerance Percent Defective is defined by the American Society for Quality Control as ". ..expressed in percent defective, the poorest quality in an individual lot that should be accepted." 2. Consumer's Risk (or 0), the risk of accepting a lot of quality equal to the LTPD, is defined by the American Society for Quality Control as "Risk, Consumer's-(J3)-For a given sampling plan, the probability of accepting a lot, when the sampling plan is applied to a submitted lot or process of a given relatively poor quality, whichever is applicable." 3. Acceptance Number means the largest number of defectives (or defects) in the sample or samples under consideraticn that will permit the acceptance of the inspection lot. 4. Acceptance Sampling means sampling inspection in which decisions are made to accept or reject product: also, the science that deals with proceduies by which decisions to accept or reject are based on the results of the inspection of samples.Note 1: The alternative "rejection" for purpose of practice the alternative may outright rejection.to acceptance is termed the definition, although in take some form other thanNote 2: In lot-by-lot sampling, acceptance and rejection relate to individual lots. In continuous sampling, acceptance and rejection relate to individual units, or to blocks of consecutive units, depending on the stated procedure. 5. Defect means an instance of a failure to meet a requirement imposed on a unit with respect to a single quality characteristic.USAEC REGULATORY GUIDES Copies of published guides may be obtained by request indicating the divisions desired to the US. Atomic Energy Commission, Washington, D.C. 20545, Regulatory Guides are issued to describe and make available to the public Attention: Director of Regulatory Standards. Comments and suggestions for methods acceptable to the AEC Regulatory staff of implementing specific parts of improvements in these guides are encouraged and should be sent to the Secretary the Commission's regulations, to delineate techniques -ed by the staff in of the Commission, US. Atomic Energy Commission, Washington. D.C. 20545. eZluating specific problemris or postulated accidents, or to provide guidance to Attention: Chief, Public Proceedings Staff. applicants. Regulatory Guides are not substitutes for regulations and compliance with them is not required. Methods end solutions different from those set out in The guides are issued in the following ten broad divisions: the guides will be acceptable if they provide a basis for the findings requisite to the iasuance or owntinuance of a permit or license by the Commission. 1. Power Reactors 6. Products 2. Resmerch and Test Reactors 7. Transportation 3. Fuels and Materials Facilities 8. Occupational Health Publiahed guides will be revised periodcically, as appropriate. to accommodate 4. Environmental and Siting 9 Antitrust Review comments and to reflect new information or experience. 5. Materials and Plant Protection 10 General 6. A Defective means a defective unit; a unit of product that contains one or more defects with respect to the quality characteristic(s) under consideration. 7. Inspection means the process of measuring, examining, testing, gaging, or otherwise comparing the unit with the applicable requirements. 8. Operating Characteristic Curve for an Acceptance Sampling Plan (OC Curve) means a curve that shows for an acceptance sampling plan the relation between the probability of acceptance and the submitted lot or process quality, whichever is applicable. Expressed another way: A curve that shows for an acceptance sampling plan the percentage of lots that may be expected to be accepted for all possible submitted lot or process qualities, whichever is applicable. 9. Sample means, in acceptance sampling, one or more units of product (or a quantity of material) drawn from a lot for purposes of inspection to reach a decision regarding acceptance of the lot. 10. Sampling, Single means sampling inspection in which the decision to accept or to reject a lot is based on the inspection of a single sample. 11. Sampling, Double means sampling inspection in which the inspection of the first sample leads to a decision to accept a lot, to reject it, or to take a second sample; the inspection of a second sample, when required, leads to a decision to accept or to reject the lot. 12. Sampling at Random, as commonly used in acceptance sampling theory, means the process of selecting sample units in such a manner that all units under consideration have the same probability of being selected. Note: Actually, equal probabilities are not necessary for random sampling-what is necessary is that the probability of selection-be ascertainable. However, the stated properties of published sampling tables are based on the assumption of random sampling with equal probabilities. An acceptable method of random selection with equal probabilities is the use of a table of random numbers in a standard manner. The sampling tables of §32.110(b) were adapted from the Dodge and Romig sampling inspection tables* which are the most commonly used tables indexed directly for LTPD (among other criteria). The consumer's risk, 3, for the Dodge and Romig tables is set at 0.10, and also is set at 0.10 for the sampling plans given in §32.11 0(b). The tables of §32.110(b) are based *H. F. Dodge and H. G. Romig, "Sampling Inspection Tables," 2nd ed., John Wiley & Sons, Inc., New York, 1959.on the simplest attribute sampling procedure that will give the required protection against acceptance of a lot of poor quality, in terms of LTPD, with a consumer's risk' of 0.10. There is no intention of optimizing efficiency of total inspection effort. To do this would require taking into account the process average quality level. A licensee may improve efficiency by selecting from the ful; set of Dodge and Romig sampling inspection tables a plan for the- designated LTPD and for his process average. Double sampling available in the Dodge and Romig tables is more efficient than single sampling and gives essentially the same protection. Typical operating characteristic curves for the sampling tables of §32.110(b) are given in Figures 1 through 8, identified by sample size n and acceptance number c. Each curve was computed for the largest lot size of the interval to which the sample applies, using the hypergeometric distribution. For any given LTPD, operating characteristic curves for sample sizes other than those plotted, computed on the same basis, would generally fall between the two typical curves shown. Values of LTPD for which tables are given should be chosen when designating the LTPD for characteristics for which sampling risks are allowahle. It should be remembered that LTPD represents the poorest quality which should rarely be accepted. The manufacturing goal should be a process whose actual process average quality level is substantially better than the LTPD.
C. REGULATORY POSITION
The acceptance sampling procedures set out §32.110 represent the minimum procedures adequately ensure conformance to requirements.in toAlthough the tables of §32.110(b) are based on attributes, variables measurements converted to attributes information would be a generally acceptable method for complying with the procedures. It is not the intent, however, to preclude a licensee from taking advantage of the more efficient methods which may be applicable to his processes, provided they afford at least equivalent quality assurance. Under § §32.15(b), 32.55(c), and 32.62(d), an application for a license or for amendment of a license may include a description of procedures proposed as alternatives to the procedures prescribed in §§32.15(aX2), 32.55(b), and 32.62(c). A variables sampling plan, or properly documented process control data, for example, might be applicable and more efficient. The licensee would be expected to show that the operating characteristic curve or confidence interval estimate for his procedure meets the required LTPD at the consumer's risk of 0.10.6.6-2I BIBLIOGRAPHY1. A. J. Duncan, "Quality Control and Industrial Statistics," 3rd ed., Irwin, Homewood, Illinois, 1965. Duncan's book presents theory and principles for analyzing and comparing various standard plans for effectiveness, efficiency, and economy. In addition, he includes material useful for designing sampling plans for optimum operation for special conditions. Duncan also covers in a similar fashion related subjects in the field of quality control, such as control charts, tests of hypotheses, and analysis of variance. 2. E. L. Grant, "Statistical Quality Control," 3rd ed., McGraw-Hill Book Company, New York, 1964. Chapters on Probability Theory and Acceptance Sampling treat principles and theory of the most commonly used acceptance sampling plans including Dodge-Romig tables and MIL-STD-105D (the current revision of military sampling by attributes). 3. J. M. Juran and F. M. Gryna, "Quality Planning and Analysis," McGraw-Hill Book Company, New York, 1970. Chapter 17 entitled "Acceptance Sampling" is a concise treatment of the practical application of sampling plans. This chapter explains the concept of sampling in relatively nontechnical terms, discusses briefly the economics of sampling, sampling risks, andsampling criteria. There is given a very abbreviated section (17-5) on theory, with reference to another source, and examples of analyses of some common sampling plans. A comparison is made of Attributes and Variable plans. Concise descriptions of the following commonly used plans are given: DodgeRomig sampling inspection tables, MIL-STD-105D, and MIL-STD-414 (military sampling by variables). The use of other information such as control chart evidence that the process is in a state of control is also briefly summarized. 4. J. M. Juran (ed.), "Quality Control Handbook," 2nd ed., McGraw-Hill Book Company, New York, 1962. Acceptance sampling plans specifically are covered in section 13 from pages 13-69 to 13-118. Material covered is about the same as the references above, but in the form and style of a manual or handbook. Section 8 entitled "Acceptance of Quality" covers such general subjects as inspection planning, interpretation of specifications, classification of quality characteristics, providing instruments, judging conformance, physical control of product, rejection of vendor-supplied product, fraud and collusion, flinching, and inadvertent shipment of defectives. This section covers practical problems and ramifications of operating the product acceptance function.6.6-3 0 0.1 0.2 0.30.40.5LOT QUALITY LEVEL, PERCENT DEFECTIVEFigure 1. -Operating characteristic curves and sampling table for Lot Tolerance Percent Defective 0.5%.0 0.25 0.5 0.75 1.0 LOT QUALITY LEVEL, PERCENT DEFECTIVE Figure 2. -Operating characteristic curves and sampling table for Lqt Tolerance Percent Defective 1.0%.6.6-41.0o 0.75 C..) z IC. 0.50 0 -J 0 "" 0.251.0 Z 0.75 IUJ U. w C-, C-,r 4 0.5 00.25 LTPD = 2% LOT SIZE 1 75 76- 100 101 -200 201 -300 301 -400 401 -600 601 -800 801 -4000 4001 -10,000 10,001 -100,000n ALL 70 85 95 1 00 105 110 115 195 200C 0 0 0 0 0 0 0 0 00 1 2 LOT QUALITY LEVEL, PERCENT DEFECTIVE Figure 3. -Operating characteristic curves and sampling table for Lot Tolerance Percent Defective 2%.LTPD = 3% LOT SIZE 1 -40 ALL 0 41 -55 40 0 56- 100 55 0 101 -200 65 0 20 1 -500 70 0 501 3000 75 0 3001 1000,00 1300 1 2 3 LOT QUALITY LEVEL, PERCENT DEFECTIVEFigure 4. -Operating characteristic curves and sampling table for Lot Tolerance Percent Defective 3%.6.6-51.0.0.75 w ci 2 c. " 0.5 0 I-.J a. 0.25n 200 C=1i = 130 C =1'I4 C" 0.75 U z CJ Ia. U U S0.50 I.-J 0 a. 0.250 1.00 1 2 3 4 5 LOT QUALITY LEVEL, PERCENT DEFECTIVE Figure 5. -Operating characteristic curves and sampling table for Lot Tolerance Percent Defective 4%.LTPD = 5% LOT SIZE ti C" 1 -30 ALL 0 31 -50 30 0 51 100 37 0 101 200 40 0 201 -300 43 0 301 400 44 0 401 2000 45 0 2001 100,000 75 11 2 3 45l.OT QUALITY LEVEL, PERCENT DEFECTIVE Figure 6. -Operating characteristic curves and sampling table for Lot Tolerance Percent Defective 5%.6.6-61? = 95." =1LTPD -4% LOT SIZE t7 C 1 -35 ALL 0 36-50 34 0 51 -100 44 0 2 101 --200 50 0 201 -2000 55 0 2001 -100,000 95 1.< 0.75 a. z LU L) L 0.5 L. 0 Iko 0.25 0.01.0S0.75 a z -L. w 4 0.50 LL 0 I-. -J o 0.25 a.0/l = 75 C'=1c =00[[ 4 (L w CL Uj LL 0 4 0 C.L 0 >,J 00 1 2 3 4 5 6 7 LOT QUALITY LEVEL, PERCENT DEFECTIVE Figure 7. -Operating characteristic curves and sampling table for Lot Tolerance Percent Defective 7%.LTPD = 10%n1 =39 c =1LOT SIZE 11 C 1 -20 ALL 0 21 -50 17 0 51 -15 0 20 0 101 -200 22 0 201 -1800 23 0 801 -100,000 r319 17.5 10LOT QUALITY LEVEL, PERCENT DEFECTIVE Figure 8. -Operating characteristic curves and sampling table for Lot Tolerance Percent Defective 10%.6.6-71.0 0.754 0. w U 0. U 0 I,.J 4 0 0.002.55 }}