108 Sampling Procedures

Quality Control and Reliability HANDBOOK (interim) H 108 SAMPLING PROCEDURES AND TABLES FOR LIFE AND RELIABILITY TESTING (Based on Exponential Distribution) 29 APRIL 1960 OFFICE OF THE ASSISTANT SECRETARY OF DEFENSE (Supply and Logistics) WASHINGTON 25, D.C.OFFICE OF THE ASSISTANT SECRETARY OF DEFENSE WASHINGTON 85, D.C SUPPLY AND LOGISTICS Sampling Procedures and Tables for Life and Reliability Testing Quality Control and Reliability H-108, is handbook (interim) was developed by the Bureau of Naval Weapons, Department of the Navy on behalf of the Office of the Assistant Secretary of Defense (Supply and Logistics). It was prepared to meet a growing need for the use of standard sampling plans for life and reliability testing Such plans may be used to demonstrate the conformance of equipments, subassemblies and component parts to established reliability requirements. ‘Suggestions for improvements of this handbook should be addressed to the Office of the Assistant Secretary of Defense (Supply and Logistics), Washington 25, D.C.FOREWORD ‘This handbook has been prepared to meet growing need for the use of standard sampling procedures and tables for life and reliability testing in Government procurement, supply, and maintenance quality control operations as well as in research and development activities where applicable. A characteristic feature of most life tests is that the observations are ordered in time to failure. If, for example, 20 radio tubes are placed on life test, and t, denotes the time when the i'th tube fails, the data occur in such a way that tS... S ty The same kind of ordered observations will occur whether the problem under consideration deals with the life of electric bulbs, the life of electronic components, the life of ball bearings, or the length of life of human beings after they are treated for a disease. ‘The examples just given all involved ordering in time. In destructive testing involving such situations as the current needed to blow a fuse, the voltage needed to break down a condenser, the force needed to rupture a physical material, the test can often be arranged in such a way that every item in the sample is subjected to precisely the samme stimulus (current, voltage, stress). If this is done, then clearly the weakest item will be observed to fail first, the second weakest next, ete. While the random variable considered mostly in this handbook is time to failure, it should be emphasized, however, that the methodology provided herein can be adopted to the testing situations mentioned above where the random variable is current, voltage, stress, ete Chapter 1 of the handbook describes general procedures and definitions of terms used in life test sampling. Chapter 2 describes specific procedures and applications of the life test sampting plans for determining conformance to established reliability requirements. An explanation of the system used in numbering the pages and paragraphs itv this handbook is appropriate, Each chapter carries its own independent sequence of page, paragraph, and table numbers, Pages are numbered as follows: the figure preceding the decimal point represents the number of the chapter and the numbers following the decimal point denote the page within the chapter. ‘Thus, page 2.16 means chapter 2, page 16. Paragraphs are numbered so that the first figure represents the number of the chapter, the uppercase letter represents the section within the chapter, and finally the paragraph number itself For example, 2A1.5 means chapter 2, section A, and paragraph 1.5, Tables and graphs are numbered with the first figure representing the number of the chapter, the uppercase letter representing the section, and finally the table or figure number itself. Thus, table 2C-5 means chapter 2, section ©, table 5, and fig. 2D-3 means chapter 2, section D, figure 3 Whenever the methodology or choice of procedures in the handbook require clarification, the user is advised to consult « qualified mathematical statistician and reference should be made to appropriate technical reports and other publications in the fieldTABLE OF CONTENTS Tine Page Foreword iit Chapter 1. INTRODUCTION AND GENERAL DEFINITIONS OF TERMS hi Chapter 2. SAMPLING PROCEDURES AND TABLES 2i Introduction 2iii vetion 2A. General Description of Life Test Sampling Plans 21 2B. Life Tests Terminated Upon Occurrence of Preassigned Number of Failures 2.26 2C. Life Tests Terminated at Preassigned Time 242 2D. Sequential Life ‘Test Sampling Plans 2.96 Appendix 2D-1 i i 2.00 vilINTRODUCTION The theory underlying the develorment of the life test sampling plans end procedures of chapter 2 of this handbook assume the measurements of the length of life are drawn from an exponential distribution, Statistical test procedures for determining the validity of Ue exponential distribution wssumption have appeared in the technical statistical journals. Professor Benjamin Epstein has published a com- prehensive paper (in two parts) on this subject in the February and Mav 1960 issues of Technometrics."* Part I of the paper contains descriptions of the mathematical and graphical procedures ar well as an extensive bibliography for reference purposes. Numerical examples illustrating the statistical procedures are included in Part IT of the paper. Epttela, “Tests for the Valldlty of the Assumption that the Underlying Distribution of Life is Exponentia,”” Technometrice, Feb. 1960 and May 1960. VitGENERAL DEFINITIONS OF LIFE AND RELIABILITY TEST TERMS IAL SCOPE 1A1.1 Purpose. Chapter 1 of this handbook provides definitions of terms required for the use of the life test sampling plans and procedures of chapter 2. 1AL2 Life Test. Life test is the process of placing the “unit of product" under a specified set of test conditions and measuring the time it takes till failure, 1AL3° Unit of Product. The unit of product is the entity of product which may be placed on life test. 1AL4 Specifying Failure, ‘The state that constitutes a failure must be specified in advance of the life test. TALS Life Test Sampling Plan. A life test sampling plan is & procedure which specifies the number of units of product from a lot which are to be tested, and the criterion for determining acceptability of the lot. 1A1.6 Life Test Terminated Upon Occurrence of Preassigned Number of Failures. Life test zampling plans whereby testing is terminated when a presssigned termination number of failures, r, occur are given in section 2B of this handbook JAL.7 Life Test Terminated at Preassigned Time. Life test sampling plans whereby testing is terminated when w preassigned termination time, 7, is reached are given in section 2C of this handbook. 1A1.8 Sequential Life Test. Sequential life test is a life test sampling plan whereby neither the number of failures nor the time required to reach x decision are fixed in advance but instead decisions depend on the accumulated results of the life test. Information on the observed time to failure are accumulated cover time and the results at any time determine the choice of one among three possible decisions: (1) the ot meets the acceptability criterion, (2) the lot does not meet the acceptability criterion, or (3) the evi- dence is insufficient for either decision (1) or (2) and the test must continue, Sequential life test sampling plans are given in section 2D of this handbook and have the advantage over the life test sampling plans mentioned in paragraphs 1A1.6 and 1A1.7 in that, for the same operating characteristic curve, the expected waiting time and the expected number of failures required to reach w decision ns to lot wecepta- bility are less for the sequential life tests, 1A19 Expected Number of Failures. The number of feilures required for decision is the number of failures that have occurred at the time the decision as to lot acceptability’ is reached. For the life test, sampling plans mentioned in paragraph 141.6, thio number of failures ia known in advance of the life test; but, for the sampling plans mentioned in paragraphs 1A1.7 and 1A1.8, this number cannot be predetermined. The expected number of failures required for decision is the average of the number of failures required for decision when life tests are conducted on a large number of samples drawn at random from the same exponential distribution. The expected number of failures can be predetermined for the sampling plans mentioned in paragraphs 1A1.6, 1A17, and 118. 1AL.10. Expected Waiting Time. ‘The waiting time required for decision is the time elapsed from the start of the life test to the time decision is reached as to lot acceptability. ‘The waiting time required for decision cannot be predetermined for any of the sampling plans mentioned in paragraph 1A1.6, 1A1.7, or 1A1.8. The expected waiting time required for decision is the average of the waiting times required for decision when life tests are conducted on a large number of samples drawn at random from the same laexponential distribution, ‘The expected waiting time can be predetermined for the sampling plans men- timed in paragraphs 1A1.6, 1AL.7, and 1A18, 1A2. LENGTH OF LIFE 1A21 Length of Life. ‘The terms “length of life’” and “time to fuilure’’inay be used interchangeably and shall denote the length of time it takes for « unit, of product to fail after being placed on life test. ‘The length of time may be expressed in any convenient time scale such ws seconds, hours, days, ete 1A2.2. MeanTimetoPailure. ‘The terms'‘mean time to fuilure’’and “mean life” m ably and shall denote the mean (or equivalently, the average) lengtle of life of i life is denoted by 1A2.8 Acceptable Mean is considered satisfactory. y beused interchange- usin the lot. Mean fife. ‘The acceptable mean life, dp, is the minimum mean time to failure which 1A2.4 Unacceptable Mean Life. The unacceptable mean life, 6, (@<é), is the mean time to failure such that lots having mean life less than or equal to 6, are considered unsatisfactory, ‘The interval between 6 and &, is a zone of indifference in which there is « progressively greater degree of dissatisfaction as the mean life decreases from 8 to 6, 143. FAILURE RATE 1A3.1 Proportion of Lot Failing Before Specified Time. The term “proportion of lot failing before specified time,” p, denotes the fraction of the lot that fails before some specified time 7, Le, exp(—T18) 1A3.2 Failure Rate During Period of Time. The “failure rate during period of time 7,” C,is given by — apf cap (11) Feit. 1A3.3 Instantaneous Failure Rate. The “instantaneous failure rate” or “hazard rate” given by- Z=\. 1A3.4 Acceptable Proportion of Lot Failing Before Specified Time. ‘The “acceptable proportion of lot failing before epecified time," po, is the maximum fraction of the lot that may fail before the specified time T and still result in the lot being considered satisfactory 1A2.5. Unacceptable Proportion of Lot Failing Before Specified Time. ‘The ‘unacceptable proportion af Jot failing before specified time," p,, (p1>Po), is the minimum fraction of the lot that may fail before time T and results in the lot being considered unsatisfactory. ‘The interval between ps and p; is a zone of indifference in which there is a progressively greater degree of dissatisfaction as the fraction of the lot failing before time 7 increases from Po to’. 1A3.6 Acceptable Failure Rate During Period of Time. The “acceptable failure rate during period of time,” Gy, is the maximum failure rate during period of time that can be considered satisfactory. 1A3.7 Unacceptable Failure Rate During Period of Time. ‘The ‘unacceptable failure rate during period of time,” G;, (G,>0), is the minimum failure rate during period of time that results in the lot being. considered unsatisfactory. ‘The interval between G, and G, is « zone of indifference in which there is a progressively greater degree of dissatisfaction as the failure rate increases from Ge to Gs 143.8 Life Test Sampling Plans Based on Failure Rat failure rates are given in section 2C, part IIT 1A4 OPERATING CHARACTERISTIC CURVES AND SAMPLING RISKS 1A4.1 Operating Characteristic Curves. ‘The operating characteristic (OC) curve of a life test sampling plan is the curve which shows the probability that e submitted lot with given mean lie would meet the ‘acceptability criterion on the basis of that sampling plan. ue 1. Life vest s pling plans which are based on1A4.2 Producer's Risk. The producer's risk, a, is the probability of rejecting lots with mean life 6. For the procedures of section 2C, part IIT, the producer's risk may also be defined as the probability of rejecting lots with acceptable proportion of lot failing before specified time, py 144.3 Consumer's Risk. ‘The consumer's risk, 8, is the probability of accepting lots with mean life @, For the procedures of section 2C, part 111, the consumer's risk may also be defined as the probability of accepting lots with unacceptable proportion of lot failing before specified time, p. 1A5 SUBMITTAL OF PRODUCT IAS. Lot. ‘The term “lot” shall mean either an “inspection lot”, i., a collection of units of product, manufactured under essentially the same conditions, from which a sample is drawn and tested to deter mine compliance with the acceptability criterion; or, @ “preproduction lot”, i.e., one or more units of product submitted prior to initiation of production for test to determine compliance with the acceptability criterion. 146 SAMPLE SELECTION 146.1 Drawing of Samples. A sample is one or more units ot product drawn at random from a lot, 1A6.2 Testing Without Replacement. Life test sampling without replacement is a life test procedure whereby failed units are not replaced 146.3 Testing With Replacement. Life test sampling with replacement is a life test procedure whereby the life test is continued with each failed unit of product replaced by @ new one, drawn at random from the same lot, as soon as the failure occurred. In the case of complex unit of product, this may be in- terpreted to mean replacement of the component which caused the failure by a new component drawn at random from the same lot of components. When the “sample sizes” are the same in both instances, the expected waiting time required for decision when testing with replacement is less than when testing without replacement. 146.4 Sample Size. The sample size, n, for a life test is the number of units of product placed on test at the start of a life test. When testing with replacement, the total number of units of product placed on test will, in general, be greater than the original sample size. ‘The sample sizes, for the life test plans of chapter 2, depend on the relative cost of placing large numbers of units of product on test and on the expected length of time the life tests must continue in order to determine acceptability of the lots. _In- creasing the sample size will, on one hand, cut the average time required to determine acceptability but on the other hand will increase the cost due to placing more units of product on test. 147 EXPONENTIAL DISTRIBUTION 147.1 Exponential Distribution With One Parameter. The density function for the exponential distribution with one parameter is given by— J9)=1Jéexp(—110) 120, 0>0 <0 ‘and the function has the following general graphical form: £410) 131A7.2 Exponential Distribution With Two Parameters. The density function for the exponential distribution with two parameters is given by~ 19,4) {6 ezpl—(t—A)/o] 12 AZO 0 elsewhere and the function has the following gencral graphical form: —+{ 4 guarantee tine £(t30,4) The quantity A is called “guarantee time” and the one parameter case is a special case of the two parameter distribution with a guarantee time of zero. 1AT$ Exponential Distribution When Number of Parameters Is Unspecified. In this handbook, whenever the term “exponential distribution” is mentioned without specific mention of the number of parameters, it shall he assumed to mean the exponential distribution with one parameter. M4CHAPTER 2 SAMPLING PROCEDURES AND TABLES (Exponential Distribution) 2INTRODUCTION ‘The theory underlying the development of the life test sampling plans of this chapter, including the operating characteristic curves, assume that the measurements of the length of life arc drawn from fan exponential distribution, 1 is important to note that the life test sampling plans of this chapter are not te he used! indise criminately, simply beeause it is possible to obtain life test data, Only after the exponential assumption is deemed reasonable should the sampling plans of this chapter be used This chapter is divided into four sections. ection A describes general procedures and description of life test sampling plans. Section B describes specific procedures and applications of sampling uated upon die occurreuce of a preassigned umber of failures and see tion C provides sampling plans when life testa are terminated at a preassigned time. Section D describes sequential life test sampling plans. Section B is divided into three parts: (1) Acceptance Procedures: (2) Expected Duration of Life Tests and Cost Considerations in Selection of Sample Sizes; and (3) Life Test Plans for Certain Speci- fied Values of a, 8, and 6/8. Section C is also divided into three parts: (1) Acceptance Procedures. (2) Life Test Plans for Certain Specified Values of a, 8, 8/9, and T/é: and (3) Life Test Plans Based on Proportion of Lot Failing Before Specified ‘Time. Section D consists of only one part Operating characteristic (OC) curves for the life test sampling plans of section 2B, part I: section 2C, part I, and section 2D are shown in table 24-2 for the corresponding sampling plans in these see= tions were matched with respect to their OC curves. ‘The oporating characteristic eurves in table 2A-2 have been computed for the life test sampling plans of section 2B, part I, but are equally applicable for the sampling plans of section 2C, part I, anud section 2D. ‘The procedures of this chapter are based on the premise that the life tests are monitored continuously. If the tests are monitored only periodically, the values obtained from the tables and curves are only approximations, plas when life cests are Le 2iliSECTION 2A GENERAL DESCRIPTION OF LIFE TEST SAMPLING PLANS 241 SCOPE 2AL.1 Purpose. Chapter-2 of this handbook establishes life test sampling plans for determining acceptability of a product when samples are drawn at random from an exponential distribution. 2AL.2 Specifying Acceptable Mean Life. Before the start of the life test, the particular value of the acceptable mean life, 4, shall he specified except when using the procedures of section 2C. part IIL 2A1.3. Specifying Unacceptable Mean Life. The particular value of the unacceptable mean life, @, hall be specified in advance of the life test when using the life test procedures of section 2B, part III, and section 2C, part IT 2AL4 Specifying Acceptable Proportion of Lot Failing Before Specified Time. The particular value, ‘Ps, of the acceptable proportion of lot failing before specified time to be used in the life test shall be specified in advance for the procedures of section 2C, part IIT 2AL.5 Specifying Unacceptable Proportion of Lot Failing Before Specified Time. The particular value, ps, af the unacceptable proportion of lot failing before specified time shall be specified in advance of the life test when using the procedures of section 2C, part ITT. 242 SAMPLING RISKS 2A2.1. Producer's Risk. The producer's risk, a, is the probability of rejecting lots with mean life For the procedures of section 2C, part III, the producer's risk may also be defined as the probability of rejecting lots with py as the acceptable proportion of lot failing before specified time. Summarized below are the various numerical values of a and the master sampling tables in which they are given. Section 2B, part I 0.01, 005, 0.10, 0.25, 0.50 2B Section 2B, part IIT Of, 08, 10, 25 28-5 Section 2C, part I 1, 08, "10, "25, 50 2c-1, 2c-2 Section 2C. part TL. 1, 08, 10, 28 20-3, 20-4 Bection 2C, part IIL 01, 05, 10 2C-8 Beetion 20, 01, (05, 10, 25, 0 2b 2A2.2 Specifying Producer's Risk. The particular value of a to be used in the life test shall be selected from among those given in paragraph 242.1 and specified in advance of the life test 2A2.3 Consumer's Risk. ‘The consumer's risk, 8, is the probability of accepting lots with mean life 4,, For the procedures of section 2C, part TIT, the consumer's ris also be defined as the probability of accepting lots with p, as the unacceptable proportion of lot failing before specified time. Summarized below are the various numerical values of 8 and the master sampling tables in which they are given. Seetion 2B, part 1 = 10. 2B1 Section 2B, part IIL : 01, 08, 10, 28 Seetion 2C, part I... io Seetion 2C, part IT (01, 05, 10, 25 Seetion 2C, part ITT 1, 08, 10 Section 2D 10 ‘The smaller the value of 2, the greater is the protection against acceptance of lots with low mean life or high failure rave 2a2A24 Specifying Consumer's Risk. The particular value of 8 to be used in the life test shall be selected from among those given in paragraph 242.8 and specified in advance of the life test. 2A3 OPERATING CHARACTERISTIC CURVES 2A3.1 Operating Characteristic Curve. ‘The operating characteristic (OC) curve of a life test sampling plan is the curve which shows the probability that @ submitted lot with given mean life would meet the acceptability criterion on the basis of that sampling plan, ‘The OC curves given in table 24-2 are equally applicable for the sampling plans of section 2B, part I; section 2C, part I: and section 2D of this chapter. Moreover, the OC curves are also equally applicable for both the sampling with and without replacement procedures. ‘The abscissas of the OC curves are expressed as the ratio 4/8 in table 2-2 so that the same set of OC curves is applicable regardless of the value of the specified acceptable mean life 6 243.2 Sampling Plan Code Designation. The life test sampling plans of section 2B, part T; section 2C, part I; and section 2D, along with their associated OC curves. are designated by code letters and numbers, ‘he sample code is given in table 24-1 and is determined by the values of a, 8, and 64/4. ‘The OC curves of all sampling plans designated by the same code pass through the two points (1, 1a) and (6/6, 8=0.10). Thus, all sampling plans which are designated by the same code offer essentially the same protection. 2A3.3 The Ratio ¢,/6, as Measure of Protection Offered by Sampling Plan. The consumer's risk Bhas been defined in parageaph 242.3 as the risk of accepting lots with mean life &. Because the OC ccurves are drawn with abscissa 6/¢y, the ratio 6/8) is also & measure of mean life which is accepted with probability 8. ‘The ratio 6/8 must be greater than zero but less than unity. If a, 8, 9 are kept constant, as 61/0, increases, the protection offered by the sampling plan against accepting lots with low mean life also increases. ‘Thus table 24-1 allows comparisons in the amount of protection offered by the various sampling plans, for in any column, the protection increases as 6/4) increases. able 24-1 Life Test Sampling Plan Code Designation Index of Life Trat Sampling Plans of seetion 2B, part I; sestion 2C, part Ij and section 20. (OC curves are given for these sampling plans in table 2A 2 oot 154 205, 246, 2m an 338 361 82 459 512 550 581 025 658 7m 745Table 24-2 Operating Characteristic Curve for Life Teat Sampling Plane of section #B, part I; tection £C. port I and ection #D 23ane epee 0 HAND UO WOIIBION, 810K ere) 347 NV3N 318V1d309V / 3417 NV3W TWALoV or 6 80 £0 vo 90 ¥o £0 zo ro ° r ° ore oz0 oro ovo oso 090 ovo 080 060 001 { iuepainbe Ayo\pe8e +10 ey peudievo.exd 10 peiou\¥se) s1804 puD evOIe Io|unbeD 0) senIND) S3YuNTIVS 40 Y3IEWNN GA3NOISSV-3ud 40 JONSYYNDDO NOdN G3LVNINYSL SLS3L 3417 YO4 S3IAUND ILSIMSLOVYVHD ONILWYSdO @-v2 31eVL B3ONV1d399V JO ALMIEVEOUdowostieep epo2 word eplwon eID s8AsnD UO AUOUOIEN “#1ON Cove) 3417 NW3W 378V1d300" / 34/7 NV3N TWALOV or fo eo LO 0 890) 660) ROBO ro ° oro oro oso ovo ovo ovo ono 080 060 0071 wubyes9-a1d Jo peipuiwie; 804 puD avo|d [oljueNbes 10) seAInD) S3UNTIVS JO Y3SEWNN G3NSISSV-3Yd 4O JONSYYNDDO NOdN G3LVNINUSL SLS3L 3417 YOd SBAYND OILSIYSLOVYVHD ONILVY3dO e-V2 F1GVL weyn fvo putes e1o snAsno BONV1d300V 4O ALMIaveoudououbreep open of ‘dupe ase seasno vo svonDion 410% ayn Ayu0 pi1BA #20 se1un9 (eve) 3417 NV3N 37@V1d300V / 3417 NV3W TWnLoV 60 vo LO yo so vo £0 zo vo ° oro ozo fo ovo oso 090 ono ogo BONVIda90V 4O ALMIGVEOUd 060 001 | wuejpsinbe (1jo})u0M8e e38 eu, peubwen.esd yy peypuje) ayse; pue suDjd JoVUeNber 20) e0AINg) S3UNTIV] JO Y3EWNN G3NDISSV-3dd 40 3INSYYNIDO NOdN QG3LVNINUSL SLS3L 3417 HOS S3AUND DILSINSLOVUVHD ONILVYSdO z-ve F1evL 20Ieiwevedse v0 we: ‘euoubreap w700 up} eyduoe a10 teesn> vO WWOLLBION :9I0N oye fi eve) 3417 NVAW 318Vid300V / 3417 NV3W IWnLov on 60 eo Lo v0 <0 ° O10 ozo oro ovo oso 090 o2'0 080 060 001 ( Weroainbe (i1oHesee ox oun peubres0.00d 4D pruD\iuse) syee4 puD euDid jo}venbes 20) seAsn9 ) S3UNTIV] JO Y3EWNN GINOISSV-3¥d JO 3ONZYYNDDO NOdN QZLVNINYSL $1631 34M YO4 S3AUND ILSINSLOVEVHO ONILVYSdO 2-v2 a16vL 3ONV1d300V 4O ALMIavEoUd‘oyoubreep op02 Word eiduoe 23D $01192 uo SUOH1O10N =210N oye fu (ere) g4I7 NV3W 37@v1d300V / 3417 NVAW Twnlov on 60 80 10 90 so bo £0 z0 10 ° oro ozo oro ovo oso ovo oo oso 060 001 Jptiuanbes 49) se4ung) S3UNTIV] 40 Y3EWNN G3NSISSV-3¥d 4O 3ON3YYNDIDO NOdN G3LVNINNSL SLS3L 3411 HOS S3AUND DILSINSLOVYVHD ONILVH3dO a-v2 31evVL LV Son BONVId390V 4O ALITIaVEOUdloruevedxe vo wei oyoubnap apo vox eiduoe (ere) 3417 NV3N 378VLd399V / 3417 NV3W TWNLov zo oro ozo oro ovo oso + 0s'0 oo ogo aes oso SHEE oo (iuojpnnbe dio ueete ep ew pe food sp peyourwies ayney pup susié jonuenbes 0) 504309) S3UNTIVS 4O Y3IEWNN G3NDISSV-3dd JO 3INSYYNDIO NOdN G3LVNINYSL SLS3L 3417 YO4 SIAYND IILSINSLOVYVHD ONILVY3dO @-v2 FTevL BONVId399V 4O ALIMIEVEOUd dwoe 210N, ‘voun jvo piysh o10 senna enous 210wououbmep ops uojé ayauos oun sensna Uo suotioioN 410N (eva) 347 NV3W 378¥1d300V / 3417 NV3W TWnLov oso one ovo o6'0 001 24 pup sucid vey sang) S3YNTIV] 40 Y3SGNNN G3NOISSY-3¥d 4O 3ON3YUNIIO NOdN G3LVNINYSL SLS3L 3417 YOd SIAYND OILSINSLOVYVHD ONILWY3dO ve aevL BONVId390V 40 ALIMIGVEOud AJoruevodee uo wo, ousubieep 0709 vo} eyfut 910 H4AIRD UO AUOILDION "8I0N woys ise (eve) 3417 NV3W 318v1d300V / 3417 NV3W WNloV z rn 60 go x0 90 so oxo 080 060 oot ‘Auretiweane oo ous peubiese-ex 1B peiouluses cite) pun cvoid fouuenbes 20) #07109) S3UNTIV] 40 YIEWNN G3NDISSV-3¥d JO 3ON3YYNDIDO NOdN O3LVNINYSL SLS3L 34I7 YOS SSAUND DILSIYSLOVYVHD ONILVU3dO @-v2 aTevL BONW1d399¥ 4O ALMIVEONd orm 212nymundee vo Woy, Burd woyoubinp epe2 Uoyd exduoe ean sotsno ve SLOHHOION :ON, soya five pii0k 0 sete seou. (ere) 3417 NVAW 37@Vid300V / 3417 NV3W TWNLOV ov oe ze ez vz oz ot a #0 v0 ° ° oro ozo oro ovo oso 090 ono 080 BONVId300V JO ALITIGVEOUd 060 007 uuses e148) puo susid jolwenbes 20, s0A109) S3UNTIVI JO Y3EWNN GANDISSV-3Yd 4O 3ON3YYNDDO NOdN G3LVNINUSL SLS3L 3417 YO4 SBAUND IILSINSLOVYYHD ONILVYsdO 2-v2 evi 213,oloubisp ef00 uo stews exo sensno wo eve) a4 NVBW 378v1d300V / 3417 NV3W TWnLov oz et 9 v1 z or 80 90 FO z0 ° ore ozo oe ovo oso o90 o20 80 oso 0071 oyuenbes 4) soAung ) S3YUNTIV] JO Y3EWNN GINSISSV-3¥d JO JINSYYNIDO NOdN G3LVNINYSL SLS3L 34M YOd SSAUND DILSINSLOVYVHD ONILVY3dO z-v2 F1avL BONVId390V JO ALITIGVEOUd 244Iovevedse uo ues, Bui du vaum Kuo w vonino os0ys vooubieap pr9 word siduos 119 s00un> ue SLONDION “2/0N Cera) 34/7 NV3N 378vid300V / 3417 NV3W TWALoV <0 oro oz0 oro ovo oso oso ozo oso BONVId399V 4O ALIMIEVEOud 060 4001 (291 pun su yuondes 30} 94209) S3YUNTIVS JO Y3EGWNN G3N9ISSY-3¥d 4O 3ON3XYNDIO NOdN QG3LVNINUSL S1S31 3417 YO4 SBAYNND DILSINALOVYVHO ONILWYadO @-ve Fav. 215“vot Ioyuevode vo wos Buiduos 1e4N A1U0 plIDK W10 seAino wHmLL oioubreep apo ubHd eyéwor! em HIAIND LO SHOLIDION “AION, (ere) 3417 NV3N 378v1d300¥ / 3417 NV3N TWNLoV a t" on 60 eo Lo v0 so ° oro oz0 oso ovo oso ogo ove ovo BONV1d309¥ 4O ALMIGvEOud 060 001 nbs fyjouvesse e10 ey pavtiss0.exd 10 peioujuze) aysey puD sue|d joilvenbex 10) s9AInD) S3YUNTIVS 4O YIGWNN O3NOISSV-3¥d 40 3INSYYNIDO NOdN G3LVNINUSL S1S3L 34 YOd S3AUND DILSINSLOVYVHD ONILVYSdO 2-v2 TTevL 246ourquie wears uw wor Bu aye (ue vovjoubinep opas void aduon ou os (°ave) 347 NW3N 378V1d300V / 3417 NV3W TWnLov vo = v0 FONVId300V 4O ALMIavEOud oo (ivzjpninbe Kjouesse oe ea ed 40 peiouiuses sy8e1 pup suid jowenbes 10) 4:09) S3UNTIVS JO Y3ENNN G3N9ISSY-3¥d 40 3ON3YNND9O NOdN G3LVNINYSL SLS3L 3417 YOS S3AYND SILSINZLIVYVHD ONILVYSdO 2-ve Fev. 2a7wm ° lweveds» ue wos, Buyduos ‘vooubisep apes uoyd syduos exe soKin9 vo SUOHIEION =010N ie pac eve) 3417 NVAN 378Vid3990V / 3417 NVAW TWNLOV £0 +010 ozo oro oro oso 090 ovo wy pevrss0-218 10 S3UNTIV] JO Y3GNNN GANSISSY-3Yd JO 3ON3YYND9O NOdN QZLVNINUSL SLS3L 3311 YOS S3AYND OILSINSLOVYYHO ONILVYSdO eve aevL ine 984 BONVId300V 4O ALITMIEVEOud 248.‘vounai votre vo wey, ooubep ope» void siduoK #10 weAIna UO HUOHLOWN :810N aya ive 91108 010 eons aneML eve) 3411 NV3W 37@VLd300V / 3417 NVAN TWNLOV eI vt et 2 " or 60 @ 40 90 $0 oro oz0 oro ovo oso 219 090 oo 080 3ONVId300V JO ALMIavaoud 060 001 bievo-4sd 4@ pevouiwiey wees peo suo}! jouyenber 42) s4Aun9) S3uNTIVs 40 Y3GWNN GANSISSV-3Yd 4O JONIYYNIDO NOdN Q3LVNIAYSL SLS3L 3417 YOd SBAYND IILSIYSLIVYVHD ONILVY3dO @-v2 a1evLuoubiap seo uold #108 #10 ¥84IND vO HUONDION “810N, "/8) 3417 NVAW 37@vid300V / 3417 NV3W TWnov ot 60 80 zo 90 oro ozo oro ovo oso 090 020 oe BONVId300V JO ALIMIEVEOud 060 001 1309} S3UNTIVa JO Y3EWNN G3NOISSV-3Yd JO 3ON3YYNDDO NOdN GaLVNINY3SL SLS3L 3417 YOd S3AUND OILSINSLOVYVHO ONILVYadO z-v2 318vLrausip jevedre vo wor) unuae ‘mays fiuo pii0e e10 seku99 a80uL [vououbitap epo2 uoyd sjéwos e1o sanin9 vo SUOHIDION “2I0N, (eve) 3417 NW3W 31@vid300v / 3417 NVAW TwnLov oso os oxo ov BONVId300V 4O ALITMIGVEOud 060 0011 (wusioninde Ayosgueen feu peubieso.014 yo peyouruse) sisa1 pun sue|é jouvenbes 20) se4un9} S3YuNTIVS JO Y3EWNN GANOISSY-3¥d 4O 3ON3YYNDDO NOdN G3LVNINYSL S1S3L 3417 YO4 SSAUND DILSINSLOVYVHD ONILVY3dO @-v2 Fev. 221‘vevoubieep e909 void afewos o10 sanino vo S10}1910N :e10N, (ere) 3417 NV3W 378Vid300V / 3417 NV3W WALov +] oro ozo oro ovo ogo 090 ozo oso oso 001 (Wusjounbs Gpo1ueRse 010 ew peubisso.0%6 0 perouwes 61804 pue suDid jouiuerbas 10) seAs79) S3YNTIVS 40 Y3IGWNN G3NDISSV-3¥d 40 JON3YYNDDO NOdN G3LVNINYSL SLS3L 341 YO4 SIAUND SILSINSLOVYVHD ONILVY3dO z-ve 31evL BONVId390V 4O ALIMIGVEOUd 2(ouuavedse uo wes, 8 ‘wouaubiep ope2 Uo}d ejéuos 10 sensn> vo SUOIIION :610N, oun Ao pIl0k e1 senna exeus, (eve) 3417 NV3W 318V1d309V / 3417 NV3N TWALoV 90 vo oro ozo oro ovo oso 090 oxo 080 FONW1d390V 40 ALITIGVEOUd 0€0 001 juenbes 405. $0100) S3YUNTIVS 40 YAEWNN GAaNDISSY-3¥d 40 3ON3YYNIDO NOdN G3LVNIWU3SL S1S3L 3411 YO4 S3AUND OILSINSLOVYVHD ONILVYSdO z-ve Fev.yuavodie 9 40s ououbisep epo> vox eyduos 10 s4ksno vo suo cae (eve) 34 NVAN 318¥1d300V / 3417 NV3W TWO eo zo + 0z0 oso ovo oso 090 oxo 080 060 01 S3UNTIVI 4O Y3EWNN G3NDISSV-3dd 40 3ONIYYNIDO NOdN G3LVNINYSL SLS3L 3417 HOS SIAUND IILSINSLOWYVHD ONILVYIdO ewe a1evL ores BONWIda99¥ 4O ALMIAVEOud 2242A4 SPECIFYING ACCEPTANCE PROCEDURES In order to completely identify the sampling plan to be used, the following must be specified for the sampling plans of section 2B, part 1 487, the lot met the acceptability criterion, Table 28-1 Master Tobl_ for Life Teste Terminated upon Occurrrnce of Preassigned Number of Failures em | cote cam | Cote om 0.106 | D1 0.288 | Et 0.693 266 | D2 as | £2 839 867 | Do. 318 Es B01 436 | Dot esa | a 918 387 | D5, or4 | eS 934 225 | D-6 70s | Fo 945 ssa | 17 on | et 9s 582 | D8. as | bs 980 so4 | D-9 760 | 9 963 c2z | D-10 73 | eto 987 esr | D1 si6 | E11 978 726 | D2 ge2 | E12 983 ipa) Das. soo | eis. 987 ma | 80 S08 | Dus 992 wai | Dae 993 #55 | Do 996 100 | An 782 sr Dog 997 Proves a a ee probability of eet ts with cane ‘Neptane etre cae ot e267). © Farevsanation othe vo, me ar 3A) 2 and tale 24-1LIFE TESTS TERMINATED UPON OCCURRENCE OF PREASSIUNED NUMBER OF FAILURES Part I EXPE ‘TED WAITING TIME OF LIFE TESTS AND COST CO! ‘TION OF SAMPLE SIZES ERATIONS IN SELEC- 285 SAMPLE SIZES ‘The operating characteristics of the life test sampling plans of section 2B, part 1, are independent of the number of units of product placed on test. ‘Thus, all tesis based on common values of the termination number rand producer's risk a are equally good and the choice of the sample size n depends only on the relative cost of placing a large number of units of product on test and on the expected waiting time required for decision. For fixed « and r, increasing n will, on one hand, cut the expected waiting time; but will, on the other hand, increase the cost due to placing more units of product on test. ‘This part of the handbook provides procedures for determining the optimum sample size based on. considerations of cost. 285.1 Expected Waiting Time. ‘The mean life of the lot and, as noted in paragraph 2B5, the size of the sample drawn from the lot affect the expected waiting time required to observe the r'th failure in a sample of size n. ‘The r'th failure is expected to occur more quickly in samples drawn from lots with low values of mean life, ‘The values of the expected waiting time divided by the mean life of the Jot, when testing without replac given in tables 2B-2(a) and 2B 2(b). Corresponding values for the testing with replacement situation are not tabled but may be calculated by dividing the termination number r by the sample size 2, Le, nent, Expected Waiting Time _r Mean Life of Lot 285.2 Relative Saving in Time by Increasing Sample Size When Testing Without Replacement. When testing without replacement, the expected waiting time required to observe the r'th failure in sample of size n,(n2r), may be obtained from table 2B-2(«) or 2B-2(b) by multiplying the tabled entry by the mean life of the lot. By dividing the expected waiting time when units of product are placed on test by that when only r units are placed on test, the mean life of the lot cancels out and the ratio, Expe axpected Waiting Time for r Failures in S Expected ‘iting Time for r Fait is # measure of the relative expected saving in time due to placmg more units of product on test, A brief table of these ratios is given in table 2B-3. 285.3 Relative Saving in Time by Increasing Sample Size When Testing With Replacement. When testing with replacement, the expected waiting time required to observe the r'th failure in a sample of size n is equal to the quantity r#/n. By dividing the expected waiting time when n units of product are placed on test by that when only r units are placed on test, the mean life of the lot cancels out and the ratio Relative Saving=r0jno=r/n 2B5.4 Relative Saving in Time by Testing With Replacement as Compared to Testing Without Replace- ment. When testing with replacement, the expected waiting time required to observe the r'th failure in a sample of size n (n2r) is equal to the quantity ré/m. When testing without replacement, this expected waiting time may be obtained from table 2B-2(a) or 2B-2(b) by multiplying the tabled entry measure of the relative expected saving in time due to using larger sample sizes,by the mean life of the lot @. By dividing these two expected waiting times, the mean life of the lot ‘eancels out and the ratio Exp. wait, time for r failures in sample of n when testing with replacement Exp. wi »placemen iple of» when testing without replacement is @ measure of the relative expected saving in time due to sampling with replacement. A brief table of these ratios is given in table 2B-4. Bromple 28-4 Saving in Time by Increasing Sample Site When Teating Without Replacement Example: Compare the average length of time needed to observe the failure of the first 2 out of 5 units of product under test with the average length of time required to observe the failure of 2 out of 2 units when testing without replacement. Solution: From table 2B-2(a), it is seen that for r=2 and n=2, the expected waiting time is 1.5000 6; and that for r=2 and n=5, the expected waiting time is 0.4500 6. Thus the relative saving in time by placing 5 units on test is 0.4500 @/1.500 @=0.300. This figure may also be obtained directly from table 2B-3. Hence, the average time required when 5 units are placed on test is 30.0 percent of the average time required when only 2 units are used. Brample 28-6 Saving én Time hy Inevensing Sample Sisr When Testing With Replacement Example: Make the sume compurivon as in exemple 2B-4 if the testing had been with replacement Solution: For r=2 and n=2, the expected waiting time is @ and that for r=2 and n=6 is r8jn=26/5 = 0.48. Thus the relative saving in time by placing 5 units on test is 0.48/@=0.4. Hence, the average time required when 5 units are placed on test is 40 percent of the average time required when only 2 units are used Brample 26-6 Saving in Time by Testing With Replacement Example: Compare the average length of time needed to observe the failure of the first 5 out of 5 units of product under test when testing with replacement with the average length of time needed when testing without replacement Solution: When testing with replacement, for r=5 and n=5, the expected waiting time is @. When testing without replacement, table 2B-2(a) or 2B-2(b) shows that the expected waiting time is 2.2833 4 ‘Thus the relative saving in time by testing with replacement is @/2.2833 @=0.438; or the average time required for a decision, by replacing failed units, is 43.8 percent of the average time required when failed units are not replaced. This figure may also be obtained directly from table 2B~4 2B5.5 Cost Considerations in Choice of Sample Size, Methods for finding the optimum sample size based on considerations of cost are given in this section. 2B5.5.1 Cost When Testing Without Replacement. The total expected cost of any of the life test plans of section 2B, part I, when testing without replacement, is given by— item where ¢; is the cost of waiting per uuit time ex is the cost of placing a unit of product on test. is the acceptable mean life is the termination number in the saunple size2B5.5.2 Optimum Sample Size When Testing Without Replacement. ‘The value of which mini mizes the total cost, as determined by the method of paragraph 2B5.5.1, is the optinun sample size. A general method of obtaining the optimum n is to use table 2B-2(a) or 2B-2(b). The snvallest is chosen such that the difference between the expected waiting time for the th failure when that number of units of product are placed on test and that when n-+1 units are placed on test, is less than the quantity 2/6. Bzample 28-7 Life Test Terminated Upon Occurrence of Preassigned Number of Failures Testing Without Replacement Calculation of Costs Ezample: Consider the case where r=10, 6)=1,000 hours, ¢,~$1_ per hour and ¢s=$100 per unit of product tested, Using the total cost formula, determine the optimum sample size if failed units are not replaced, Solution: Using the formula of paragraph 2B5.5.1, the costs for various values of 1» are Ered cot Coens 1029291000, S09 30200110010 12 1608 a0 mt 2646 1168 2508 151035 1% 9a 7 BT ‘The optimum sample size is thus n=16. Example 28-8 Life Text Terminated Upon Occurrence of Preassigned Number of Failures Tealing Without Replacement Obtaining Optimum Sample Size With Expected Waiting Time Example: Use table 2B-2(a) to determine the optimum sample size for the problem of example 2B-7. Solution: The quantity e,/e.8y is equal to 0.1 and from table 2B-2(a), the expected waiting times are— ire 10 2.9200 20199 0. goer 20190 60s2 0.4167 12 1.60821. 3488 0.2568 1 Lsa68 1 L682 16210889 15 Loss 0.9307 16 0.93070. 8467 7 a7 0.7773 en ee 2B5.5.3 Cost When Testing With Replacement, The total expected cost of any of the life test plans of section 2B, part I, when testing with replacement, is given eh Ftelntr—1) where ¢, is the cost of waiting per unit time 2 is the cost of placing a unit of product on test % is the acceptable mean life is the termination number nis the sample size. 2a12B.5.4 Optimum Sample Size When Testing With Replacement. ‘The value of n which minimizes the total cost, as determined by the method of paragraph 2B5.5.3, is the optimum sample size, Tn general, the optimum n for the case of testing with replacement is the integer nearest to fevBor | 1 fea Example 20-9 Life Test Terminated Upon Occurrence of Preassigned Number of Pailuree Testing With Replacement Caleulation of Cost Example: Consider the problem of example 2B-7, i.e., r=10, 8=1,000, ¢=81, and e:=$100. Using the total cost formula, determine the optimum sample size if failed units were replaced. Solution: Using the formula of paragraph 2B5.5.3, the costs for various values of m are 9 nn 1809 -2art 10 1000 1900-2900 " 900 20002909 ‘The optimum sample size is thus n= 10. Fzample 8-10 Life Test Terminated Upon Occurrence of Preaasigned Number of Failures Testing With Replacement Obtaining Optimum Sample Size by Formula Example: Use the method of paragraph 2B5.5.4 to deters of example 2B-9 Solution: The integer nearest to— ne the optimum sample size for the problem. am ny m2 4 is 10. This is the optimum sample size as was seen in example 2B-9. LIFE TES 'S TERMINATED UPON OCCURRENCE OF PREASSIGNED NUMBER OF FAILURES TESTING WITHOUT REPLACEMEN Table 28-210) Expected Waiting Time Expected waiting time for r failures in a sample of Values of Mean life of lot In the table r=1(1)n and n=1(1)20(6)30(10) 100 0.5000 0.3333 0.25000. 20000. 1087 1.5000 OBIS} ORAS 0.45000. 3087 Lau 1.0803 7BRE 0.6167 20813 1.28350. 9500 2 2833 1. 4800 24500 2320, 3005 0. 5005 0. 7505 1.0929 1. 592 2 5028 Tabie 28-20) oor 0. 8a o.2n7 0, 5226 0, 4226 0.8016 0. 9682 1, 1682 14182 0.2962 | os 1.200 0, 0667 0.1381 0, 2150 0, 2084 0, 3803 0. 4803 0, 600 0, 7254 0. 8082 1.0339 1739 Tass Laie 2.3182 3.3182 6. ovo 1908 0. 8020 0.4370 0. 5699 6,785, 0.4355, ksi 20199 10190 0.0588 0.1213 1880 0.2505 0. 5363 0. n07 0. 5106 0.6106 ora7 0, 8167 1, 9896, 1. 1362 1.3562 1.6062 14396 24390 8.43060 8 o o 0 0 ° 0. 0. 0100 0.0817 0.1251 0.1708 02182 0 2682 0.8209 0250 506 0770. 18 eu 1807, 0. 0x88 0. 0678 0.1408, 0.1790 0.2190 0 2007 Table 2B-2(a)—Continued 0.sse2 0.4176 0.4935 0.8770 0, 6688 0. 544 0.6358 0727 07901 un 2 1% 0.3514 0.3871 oa 0, 4826, 0, 5026, 0, 5443 0.887 0 6332 0.6808 234 a 2 2 % 28 2 2 20 30 west |) 18 oso a7 10089 | 18 ia | a9 isis | 20 | oss70 | 2 0.9870 |) 23 Lost | 24 ami | 35 13660 | 9326 | 17326 2111660 2 Lam 2s 4 3021 2 1580 2% LTT 2% LauT 2824950 292.9950 708) at ea) 32 a0 | 33 sore | as 60s | 35 o220 | 36 oes | 37 1733 | “38 asst | 39 ive | 0 Lon 17644 2.0077 2.5077 3.5077 1. 9826 23160 2 8160 3.1004 13 16 "7 1s 0.0167 0.0336 0.0509 0.0084 0.0863 0. 1044 Table 28-2(a)—Continued 0.3246, 0.924 3810 0. $104 0.4407 0.1720 0. 5082 0.5376 0.5720 0.0077 0 6448 0. e882 0.7232 235 0.7649 0.8085 0 ssa 0.9015 0.9515, 1.0507 18s 1 1810 | 1 | & | | i i vasa 1 1848 12403 1 2001 13616 1 4284 1 ays 1. 6000 17509 1309 1, 9620 2.0870 2 23065 25965 2 8405 3.1799 8, 6790 4s 6799 1.5702 1 9083 2 042 22159 24159 2 6659 2. 9902 3.4992 eric)0 u ir ub im u R u 15 16 " 18 0 20 ons 028 ost oz 204 186 140 208 202 ara aus ois 0252 80 uw 20 2 a Table 28-2(a)—-Continued 0 0 o 6 0 0 0 0 0 0 o 0. 740 40 00 28 soit Bnet sis sists 58 7146 023 3192 aos 4540 ang 000 74 407 $609 4863 5067 5275, 5488 5705 S028 cas? 625 600 ; 7 lens \ bs 1 feu | 48 ogo | ig [2 tae eG s2 kas | i 2 feign b zt 5146 51 206 LW 1150 2400) 182 5495 795 05 sie sus 47 704 a Sui oars ins a2: 58 1865 276 376, 5155 x21 x 121 9655, 55,Table 28-2(a)-~Continued n=90 i aout | 2 © soar | 47 i 2 toma | as oa | ae : 3 oor | amar | ts 4 Gow: | aan | oo : © come | os caus | se t tomas | 30 wana | ga 1 8 dome | at aaioe | Se 1 oon | a oates | as i 1 oat ff a lasso | 2 no oww fl s oa | ot 2 ir oi | a5 0.4890 |! 58. 22 1 Gass | ae ) 3 Bs Wotan | ar |) 2 1% uae | a a 2 ate | ae | 2 7 ower | 40 & 2 18 ois || aL ot a Ome | & 3 ~ooaer | a cs : naa fa | © : oan | as | ts Boa || .o aan a 2 2008 | 51 0.7082 | wang a 0.3170 | 52 0.7206 | bass 28 0.3266 | 53 ata | 1. 4966 20 7 1 5420 30 1,890 31 16306 32 1 6928 33 Lar 4 18068 35 1. 8691 36 19358 37 2007 28 2 ogee 39 2.1675, 0 22584 a 2 se 2 2 805 a 2.5045, 4 21974 4% 2 9040 46 3. 1040 ” 3540 8 3.0874 3 ‘sista 50 5.1874LIFE TESTS TERMINATED UPON OCCURRENCE OF PREASSIGNED NUMBER OF FAILURES. TESTING WITHOUT REPLACEMENT Table 28-216) Expected Waiting Time Expected waiting time for r failures in a sample of m ae “Meno life of lot In the table += 1(1)10(5)80(10}80/25) 100 and =r, seth k= 1(1)10(10)20 aj) Lo same | 0. 3333 | 0. 2500] 0.2000 0.1428) 0. 1250 0. 1000 2] 4 3833 | aera | 2111 nas |. 1292 1026 3] a e167 maa | 2180 1503 || 06 toss 5] 2 6456, 2182, 1518, 1083 | 6] 2 6532, 2190 eee 1044 tle se? 2106 125 | 13a ros 8] 2 520 2 2200 tsar | 132 toa? a/ 2 fest 204 ton | 11325 toa? wo} 2 688 | 2207 1530-1326 1048 is | 8 e708 | 26 tae | an 100 | 3 08 ze tae | 13 ros! | | 3 cas> | a0z2 goa | listt | lisar | “tase vost 30, 6849, 2228 1818 | 1598]. 1332 1052, 0 60 mas | isis | isan) rat ese 30 tie | zee | ts20 | “150 | “1338 tose i) to) 6808 ves | list | “isso | rae ros | roo | 8.1874 | | sn07 mao | gaz | i500} 33a 1053 | | I | Table #13 Fxpected Relative Saving in Time by Increasing Sample Size 02 0.20 39.30 50 8 ’ a “Te vals abe Mee ae theater pee lig Ue oe aur Sale oF Foray sven valu of these the tbl se the ae the tite expected saving In tne, “Thas nes Jog te camp te ml eae thereto wating Une eg or desonTable 28-4 Expected Relative Saving in Time by Testing with Replacement "shal ote ht the me sae hylan nes at's vine Ma LIFE TESTS TERMINATED UPON OCCURRENCE OF PREASSIGNED NUMBER OF FAILURES, Part UL LIFE TEST SAMPLING PLANS FOR CERTAIN SPECIFIED VALUES OF a, 8, AND. ®/@ 2B6 DESIGNING LIFE TEST SAMPLING PLANS Alife test sampling plan may be designed so that its OC curve meets the following preseribed conditions: (1) if @=6, then the probability of the lot meeting the acceptability eviterion is 1—a, and (2) if @, then the probability of the lot meeting the acceptability criterion is less than or equal t0 8. This part of the handbook, which may be considered an extension of section 2B, purt 1, provides procedures for obtaining values of the termination number r and the acceptability constant ( when certain selected values of a, 8, and 8/8 are specified. When other values of a, 8, and 6,/@ than those provided in this part of the handbook are needed, refer to section 2B, part I to determine whether one of the life test sampling plans given in that section is applicable 2B6.1 Life Test Sampling Plans. From table 2B-5, values of the termination number r and the acceptability constant may be obtained for values of a=0.01, 0.05, 0.10, and 0.25; 8=-0.01, 0.05, 0.10, and 0.25; and 4,(@=2/3, 1/2, 1/8, 1/5, and 1/10, The value of r is obtained directly from table 2B-5 but the acceptability constant Cis obtained by snultiplying the tabled entey by the acceptable mean life 239Paumple 20-11 Life Teste Terminated Upon Occurrence of Preassigned Number of Pailures Example: Find a life test sampling plan which possesses the following OC curve: If the menn life ix 4-=900 hours, the lot is accepted with probability 0.95; the mean life is 4,300 hours, itis accepted with probability approximately equal w 0.10. Solution: In this example, 6,/%= 1/3, @~0.05, and 6=0.10. Looking in table 2B-5, the termination C= (C185) -900(0.4976)~448 are obtained, In words: Stop life test ‘of lot miean life & « is greater than or equal to 448, the lot is acceptable; otherwise, the lot is not accept= able. 286.2 Expansion of Table 2B 5 for Values of 6./@, Greater Than 2/3. Approximate valurs of the termination number r and the acceptability constant C may be obtained to supplement those given in table 2B-5 for values of 8/0, greater than 2/3 provided the same values of a and 8 as given in Table 2B-5 are specified Compute nuinber r= and acceptability const reed. If the ostinato Place 8 or more units of product on Lost ng flor 8 failures have 0 Ke (0u6.) (G8) 1 and where values of Ay and Kis are tabled below aors — Kyor Ks 0.01 326 0.05 1.645 0.10 1. 282 0.25 0.674 Reample 20-12 Life Tests Terminated Upon Occurrence of Preaanigned Number of Pastures Example: Find the appropriate values of the termination number r and the acceptability constant C for the ease where the acceptable niean le #110 hours, the unacceptable mean life = 100 hours, the producer's risk «0.05, and the consunier’s risk B~0.10 Solution: From the formulas of paragraph 2B6.2 1.2824 (1.1) (1.645))"_ 9. Geter) one *Valure obtained from tables uf the cumulative Normal Distribution 240LIFE TESTS TERMINATED UPON OCCURRENCE OF PREASSIGNED NUMBER OF FAILURES Table 28-5 Life Tet Sampling Plans for Specified a, 8, and 0, aa || 136 can | tor 7 | 3 t02 | oo 728 va | 45 6 | 35 6a | go" 625 22 572 ie | is saa | ase 498 | a3 469 | toe ats 1B oe 30 | oe 363 |e 333 | 56 ino || s+ 256 | 4 1206 | e206 | eS a | os ons7 | or wos | 58 t80 | 88 Ta va | alma | 2 ams | 19.655 | 13800 i | is 1502 | toe leas | ue ayy |e aa us |) 7 la | se cam | ae cae | ae “ae wo | oe ime | sem | oe ca |e oi 2 | 52. 908 ame | 23.853 | 12708 1 laa yaa | ee ize | se ler 13 7.1726 os | ae ose | at C48 us, a 576 asi | 2 cas | e288, ino | 2 a8 aat| te lame |e S28, [Nae_A compete pet of OC curves not provide forthe sampling plan of th abie_ For those emp pls marke, he appropriate OC cures ‘tale 242 uy be sed by String, oe able 2, the eae code dagen corresponding othe same values of ud or the sapling ans 2aLSECTION 2C LIFE TESTS TERMINATED AT PREASSIGNED TIME Part 1 ACCEPTANCE PROCEDURES C1 LIFE TEST SAMPLING PLANS ‘This part of the handbook describes the procedures for use with life tests that are terminated at a specified time or upon the occurrence of a specified number of failures, if this number is reached before the specified time. ‘Two procedures are given: (1) @ procedure when testing without replacement, and (2) another procedure when testing with replacement. 2CI.1 Use of Life Test Sampling Plans. ‘To determine whether the lot mects the acceptability criterion with respect to average length of life, the applicable sampling plan shall be used in accordance with the provisions of section 2A, General Description of Life Test Plans, and those in this part of the handbook. 2C1.2 Drawing of Samples. All Samples shall be drawn in accordance with paragraph 1A6. 2¢2 SELECTING THE LIFE TEST PLAN WHEN SAMPLING WITHOUT REPLACEMENT 2C2.1 Master Sampling Table. ‘The master sampling table for the life test sampling plans of this part of the handbook is table 2-1 2C22 Obtaining the Sampling Plan. ‘The life test sampling plan consists of a tern sample size n, and an associated termination time T. 2C2.2.1 Termination Number. The termination number r shall be selected from among those given in table 2C-1 and specified prior to the initiation of the life test. ‘The choice of this number shall be dependent on the degree of protection desired against acceptance of material with unacceptable mean life. The larger the termination number, the larger is the ratio 6/8 and, as mentioned in paragraph 2A3.3, the greater is the assurance against accepting material with an unacceptable mean life. 202.22 Sample Size. ‘The choice of the sample size, as explained in paragraph 1A6.4, is dependent on the relative cast of placing a large number of units of product on test and on the expected waiting required for a decision. ‘The sample size shall be selected, with this factor in mind, from one of the following multiples of the termination number: 2r, 37, 47, 5r, Gr, Tr, 8r, Or, 107, and 207, tion number r, & 202.23 Termination Time. ‘The termination time 7, corresponding to the applicable tc number r, producer's risk a, and sample size n, is obtained from the master table by multipl tabled entry by the acceptable mean life 8 2C3. LOT ACCEPTABILITY PROCEDURES WHEN TESTING WITHOUT REPLACEMENT 2C3.1 Acceptability Criterion. ‘The acceptability of a lot with respect to a life test from this part of the handbook shall be judged by the time required for the rth failure to occur in a sample of sine n_ Compare the time of the occurrence of the r’th failure with the termination time 7, fnentioned in paragraph 20.2.2. If the 'th failure occurs prior to time 1. the lot is considered to have failed to meet the acceptability nation ing thecriterion; if the r'th failure still has not occurred by time 7, the lot is considered to have met the acceptability criterion 2C4 SELECTING THE LIFE TEST PLAN WHEN TESTING WITH REPLACEMENT 2C4.1_ Master Sampling Table. ‘The master sampling table for the life test sampling pla ment is table 20-2. 2C4.2 Obtaining the Sampling Plan. ‘The truncated life test sampling plan consists of a termination humber 7, a sample size m, and an associated termination time 7. with replace- 2C4.2.1_ Termination Number. ‘The termination number r shall be selected from among those given in table 2C-2 and specified prior to the initiation of the life test. ‘The choice of this number shall be dependent on the degree of protection desired against acceptance of material with unacceptable mean life ‘The larger the termination number, the larger is the ratio 64/@, and, as mentioned in paragraph 2A3.3, the greater is the assurance against accepting material with an unacceptable mean life 204.22 Sample Size. The choice of the sumple sie, as explained in paragraph 1404, is dependent on the relative cost of placing a large number of units of product on test and on the expected waiting time required for a decision. The sample size shall be selected, with this factor in mind, from one of the following snultiples of the termination number: 2r, 3r. 4r. 5r, 6r. 77, 8r, 97, 10r, and 20r 2C4.2.3 Termination Time. ‘The termination time 7, corresponding to the applicable termination number r, producer's risk a, and sample size n, is obtained from the master table by multiplying the tabled entry by the acceptable mean life 6. 2C5. LOT ACCEPTABILITY PROCEDURES WHEN TESTING WITH REPLACEMENT 2C5.1 Acceplability Criterion. ‘The acceptability of @ lot with respect to a life test shall be judged by the time required for the r'th failure to occur in a sample of size n. Compare the time of the oceur- rence of the r'th failure with the termination time 7, mentioned in paragraph 2C4.2.3. If the »'th failure occurs prior to time T, the lot is considered to have failed to meet the acceptability criterion; if the r'th failure still has not occurred by time 7, the lot is considered to have met the acceptability criterion, Brample 20-1 Isfe Test Terminated at Preassigned Time Te ng Without Replacement crample: Find a life vest sampling plan without replacement which will accept a lot having an aecept- able mean life of 1,000 hours with probability 0.90. ‘The experiment is to be stopped on the occurrence of the fifth failure and 10 sinits af product are ta he placed on test Solution: In the notation of this handbook, @=1,000, «0.10, r=5, and n=10—2r, From table 2c), T= A T/A.) =1,000( 314) =214 Tn wards, secept the lot if the 5th failure has not yet occurred by 314 hours and reject the lot if the Sth failure occurs before 314 hours have elapsed. Remark 1: The code designation of the above life test sampling plan is obtained from table 2C-1(c) as C5, From table 24-2, the probability of accepting a lot with mean life of. say, 500 hours may be obtained by finding the ordinate of the OC curve labeled C-5 at the point where the abscissa 4/0,= 500/1,000=0.5. ‘The probability of acceptance is seen to be equal to 0.47 Remark 2: In this example, if the termination number had been selected as 6 rather than 5, the probability of accepting a lot with mean life of 500 hours is obtained from the OC curve labeled C-6.. ‘The probability is seen to equal 0.41. This illustrates the remark made in paragraph 22.2.1 that the larger the termination number, the higher the probability of rejecting lots with unacceptable mean life. 243Bsample 80-8 Life Tost Terminated at Preastiqned Time Testing With Keplacement Example: In the problem of example 2C- been replaced. Solution: From table 2C-2(c), T=64(7/#,) = 1,000(.243)=243 hours. In words, accept the lot if the Sth failure has not occurred by 243 hours and reject the lot if the 5th failure occurs before 243 hours have elapsed. ‘Remark: The termination time, when sampling with replacement in this example is 243 hours as compared to 314 hours when sampling without replacement. This illustrates the fact that the expected waiting time for a decision as to lot acceptability is lessened by testing with replacement. find the termination time if the failed units of product had Table #C-1(0) Master Table for Life Teste Terminbled af Preassigned Time—Testing Without Replacement Values of 7M, for «0.01 cote | or ee Ma 1] 0.005 0.003 0.003 0.002 0.002 0.001 AO 001.0010. 0005 re 2| 06 027,020,018. 013011010008 008 «00K rey 3| ‘oa 1065 1040 ost ome oma orp on? 015007 ret 4] list lore Los7 1048037 ast o27 one loon as | 115 1100 107256 ous a9 ase 030 aaron ss 6] 11981117 one 1068054048 090035 Last LOS M7 7] img ligt foe Lom lop oh ota Leap Lang. ot re 8] (zn 11431103080 065055048042 gs 019 a“ 9] 1260 1185 1110088070080 052 0480100 10 ro] 1276 168111701075 063055048 04s ont wu 1s| las7La99 1421101000 078086058052 025 12 go] lar7 azz 1158123101085 07k 005058028, A138 25| 400 239 L170 1132 S108 S01 1079070082090 aM 30 | 1428 352 S179 1138 140061088074 088032 ws 4o| i460 270 Sa92 149 Laz 1103089 Lovo 070034 me so] l4e2 1263 oon 1458 lage 08 00k 108g 074 L036. mar | 1818 1304 ae 3671137116100 08s 079039 ais | 100] 184013695 Lame 1421201040202 00 or estaaton ote od pur 243. andthe 241Table #C-1(6) Master Table for Life Tests Terminated at Preastigned Time—Testing Without Replacement Values of 17, for a=0.05 cote | + = > » « ” . » * * a me BI 1] 0.028 a.017 0.013 0.010 0.009 0.007 0.008 0.006 0.0050. 008 Ba 2) 104 0850480383108 03020018 000 Ba 3] 118 108 1075-058 «048 0a 1038001028 ne Ba | fai isa os | ore on ot, ous 0m re on Bs s| ase lass i110 Los 0710801052 lous nt 090 Be iS oe ie fica 00s ots ce, fer 01 eon BT 7| 1300 a5 11321103, 0as L072 10620551008 BS | 1390 197141110000 076 066058052085 Bo 9] 13468 207) 4815 one a0 108906105507 B10 ro] 363 12161154 = 1190 098) 10s, 072 ose 1057) on Bu 16) 7 9 ie Mig oe fed ors) oes ord Buz 20| 1451 266 1801147120102 08807807084 Ba3 25] cats lam 9p ase 1128107093 ome ors 08 Bu ey a BAS | 6 mo sas eas) it moe can) Baw | ere ain es ra ea iso) oer ons on) Bar 75| see 1331235 tz 149126. 100096080 BB woo] 1581.30 2az 1187158130112 099-088 03 or xpanaion of th dew pe 243-2404 ble 2A Table #C-1(0) Master Toble for Life Teste Terminated at Preosrigned Time—Testing Without Replacement Values of Tit, for a= 0.10 cote | or ‘ample oe co 1] 0.053 0.035 0.025 0.021 0.018 0.015 0.013 O.012 OO11 0.005 cz 2] 155.008.7156 48 39k 080 027 18 ca 3{ 1276 1199 301 07908505854 OUD aR COD 4 4| lar lies 1121 005 ove oss 08r 051 us cs s| is 1891138108087 074064 «057-051 025 ce 6{ 13463 208 1471115084080 069 061055027 co 7| 1306/29 38711221100 085074065088 S028 ca n| ‘an “ann tee aR 108 ono 077068 | 061 030 co o| \aa zp amt tas tp 002 090 fort oes Coat C10 yo] ae za7 147613712085 S082 0730508 cu is} fees sis, ice en ia; 1061 0 sl! ore, ans C12 20| L404 1201207 tgk 1322007085078. 07 C13 | 1818 1303 31846713716 100089 079.080 ou go} {e900 ais ame stra mk te. 108 OL 040 cs 4o| lesz 13m aao ames aa 107 ee 8 C16 ole eer fan eae ies ieee iar 00ers ere car 75| 1600 1348245 1190115818214 101 090. 44 cas yoo} l6oe 1384251195 L1so a8 a7 103 one 4S ‘er eptanation ote ee, we par 2432 nd ie 2ATable #-1(4) Master Table for Life Tests Terminated at Preassigned time—Testing Without Replacement Values of 11, for a=0.25 = nore ba 1 0058 00s oon om oo om oon Dal lage te Com Can Cues ose oes ak | tele tetas eee ose ak ba ft a. — errr Dae | ce tas ooo ons la De |aa aan a. | te ie ooo sr ores be | 8 ie cut tor ces Sapa Dao | so] (sr car aw in io a. —rrs~—rS—CCiCKSCszsSONSS a Bu | m| [io as Oho lier lis te late os Dus | ml ‘oer ise Se lin iso lige la aor opo ane Fen OR ee ee Bas | a ‘on las le tea ctor tis Tiana ole Dis | 50 620 364 258 201 164 189 120 106, 095 046 errr Das | wo) lee cat lt bee MD oe lew Treated eo wp ad wa obi 40-10 Mame Tel for Life Tes Terminated at PreainedTime-Teting Wid Replacement Values of Ti for a=0.80 | ise er—“siOis—sisisiCi*sSrsSCss—s=sCé=édzsécCCt=z es 1] os oa ous on one ao an 0.089.038 re A a | 5 a, ... oo ous Ee i) Sci Bs cian tos lon B BE reco coeaarcnt@ fant bust er use cu oes = ,rrrr—”—r=——~—=‘=‘=—“COSCCCOSOS ooo ons ES 7) co areas amps tae 10 loa eer to low be | 3] len ie lag ae im it le ir low a rrC~—r..—C—COirsCSCizSMSCMC _ | i | | ‘eo lau at la cite it ciar oso Ce rrrrr—™rsSSC—t—t™*—“EEU ost eu | aol ‘ev caw oss tao ise igs ost Eis | | est 400 ose ota) sige ost at re ee ee ee) a eFC ost Eis | so) ‘eso cy ose ae Sige Ie tas saTable #C-£(0) ‘Master Table for Life Tests Terminated at Preassigned Time—Testing With Replacement Values of 14 for | — Sample ne Ad 1] 0.005 0.003 0.003, 0.002 0.001 0.001 0.001 0.0010. 0005, Aa Bi ost fous 010) 012-011-009 008007 « 004 At 3) 073 048-036 (oa oa ioe oie cis) fon A 4] [108 069 S051 od ae om fons) feat 010 AB 5] 1128 1085 Lone a iy oa sos ot a6 8] 149099 L074 050 1043037083030 OS at Tie 1s ons) 055 048042037033. 017 a Ol at 061 0520454003808 a 9] 195 130-097 065 05604943039 OO » 10] ‘aor 118.108 060059 052d oak » 13) 249 S168 135 083 © 071.062.055.050. 025 zo] amt 1185180 92 0780690625508 25| aor toga 099 108507086059 | 090 30] 312208186 104/089 07806906208 40) 33522387 2 1096 0ge— 1074067 L033, so] 13:0 mae 178 17.100, 088078070035 | 376 250 | 188 125 10709083075 088, roo} 391261196 130/112 098087078 039 or expacatin of he nde, par 2422 ad ale 2A Table 20-00) Master Table for Life Teste Terminated at Preastigned Time—Testing With Replacement Values of 774, for a=0.05, fal ple Ba 1| 002 0017 0.013 a010 0.009 0.007 0.000 0.000 0.005 a. 008 Be 2] 089.059 Ok 035080025022 2008. 008 Bs | 5] 1136 op 068 1055 045038034080 oa? Oe pa | 4] lam lie 10851068 1057 Logos ass ask 017 BS | 5 197 13 009 078 066 056 O49 re 39 BS | las 14s 109 087073062054 O48 Oe Br | 7] 235186117 lod 1078067089052 047 Bs | 269 1661281100083 Sort 082085050 pe | 9) lam lame 1130 S108 087 078065058082 B10 10} lat tat 1138 1109, 1080 1078068060054 Bu 15| 1306 205 ase 123,103,088 L077 08808 Bu? 20] sm zak 168153 110005083074 008 pis | a5| 348 a32 lava 1139, ite 1098087077070 Bay | 30] 360 1240 180141190108 090g 072 Bis | 40/1377 252 1891811181108 0940075 BAe | 50/130 260 119514861130 N97 087 078 Bar Ms | ae ars ek age ise sige ica, oon oes, pis | 100] az 1280210168 1401120105008 08K For exponton ofthe code orp 248 Tanda 3 2aTTable 2C-2(0) Master Table fr Life Tests Terminated at Preassigned Time—Testing With Replacement Values of 1/4, for a=0.10 co | + a C1 | 1] 0053 0.035 0.02 021 aos 0015 0.013 0012 oon 0.005 3 tes) ome en on! on a | oss oot on | is cs) 8) list ize one S073 052 Lous on loa? one ce | 4) cats 145 S100 0877s 082055 Lov Lons on cs S| m3 16212 Lon7 os L070 L061 a5 0d) one ce Oo cr Sl eave aes ao un oon ford ome ort ose 08s ce a ct od ae oT 1088 073085058020 co | a oki tol een ome, ce oe, ta ero | wo} lait 207 ase tae 0h on 078050 asa oat Gu | a5 ( 1383 gag azz sr ane 008086 076059034 cz zo} 36a ze fasz as lan 408 op ont 107s aa. cas 25 | la7t pet 4g8 ast 1126108008 ose 07s 038 ou zo] :ae7 lap asa ass aap mn opr 006 Lore Loa cs 40} 402 1268 20161 sk tts 10008908000 cos ol ee ars am isk ey fn fog ouel fo on 7s} 428 13a lata amy as La2a 10700508608 18 | 100] 437 zt ais a7 ae 1251097087 ue i or ration of vot we par 248 Bund ale Table £0-£(4) Master Table for Life Tests Terminated at Preossigned Time—Testing With Replacement Values of Tt for a=0.25, pa 1] 0.148 0.090 0.072 0.038.088 0.081 0.030.002 ees we D2 2] 2801602130 .096 080 06906) 5} MBO Da 3] 88 lise tas 118098 082072585802 D4 al sir eer tol ore am) oat Tons Ds 5) 337 1235 lags 1a 06 08s 078067034 D6 eo a ia 100 “088 07k 70-038 7) 136s 2a2 lagz las, fos Soot os 073 038 8) 37224886149 10s os 0gs 07k 057 9) 1380283 12 too ons oss 076038 10} 138658 155 no Lor 088077000 ) as} 408 laze 163 ny lien opt ope Loan Poe] au am 168 120 5105 ops ose 082 25 | 1429286 12 123 1107005086043 30} 1435 “201 74 125 1109007108704 40; 445206, 178 127 amon 089 “ona | 50) 481300 180 129° 11310009004 jm] 40 L307 184 118 a02Lovz 048 100) 485310 186 133 116403008047 ‘Fer eplaaton athe ee, or par 243 and abe 2A 248.Table #C-#(0) Master Table for Life Tests Terminated at Preassigned Pime—Testing With Replacement Values of 17% for a:=0.50 = | a = — } 1 0.347 0231 0173) 0.139 0.099 0.087 «0.0770. 069. 0.035 [a] “to 80 ato tes vos om ase on 10 | 483 322 a2 193 121 107 097 048 ar 498 332 249, 199 142, 124 un 100, 050 too] 4081 esa LIFE TESTS TERMINATED AT PREASSIGNED TIME Part II LIFE TEST SAMPLING PLANS FOR SPECIFIED a, 8, 6/6, AND T/@ 2C6 DESIGNING LIFE TEST SAMPLING PLANS, AA life test sampling plan may be designed so that its OC curve meets the following preseribed conditions: (1) if @=66, then the probability of the lot meeting the acceptability criterion is 1—a, and (2) if 6=4, then the probability of the lot meeting the acceptability criterion is approximately equal to B This part of the handbook provides procedures for obtaining values of the termination number and the sample size when certain selected values of a, 8, 8/6, and T/. are specified. This part of the handbook may be considered an extension of section 2C;, part 1, so that when values of a, 8, 8/8, and 7/8, other than those provided here are needed, refer to section 2C, part I, to determine whether one of the sampling. plans given in that section is applicable. Moreover, if the desired value of 7/8, is not given in this part, of the handbook, it may be noted that T'is usually an upper limit while 6 is @ lower limit, ‘Thus, if no sampling plan is given for the desired value of 7/%, the sampling plan for the next lower value of Tit. which is given, may be used. 2C6.1 Life Test Plans When Testing Without Replacement. From table 2C’-3, values of the termination number 7 and the sample size n may be obtaified when testing without replacement for values of #=0.01, 0.05, 0.10, and 0,25; 0.01, 0.05, 0.10, and 0.25; @/@=2/3, 1/2, 1/3, 1/5, and 1/10; and Tey 1/3, 1/5, 1/10, and 1/20. The values v 64/4 and T/@ shall be computed and for the appropriate values, of and 8, values of r and n shall then be obtained from table 20-3 249Example £0-3 Life Test Terminated at Preassigned Time Testing Without Replacement Ksample: Find a nonrepincement life test sampling plan which is not to exceed 500 hours and which will accept x Jot with acceptable mean life of 10,000 hours at least 90 percent of the time but will reject a Tot with unacceptable mean life of 2,000 hours about 90 percent of the time. Sobution: In this case T= 500, 8,~ 10,000, 6, =2,000, «=0.10, and B=0.10. Hence, 6/8 Ti8=1/20. Looking in table 20-3 under a=0.10, B=0.10, @/8==1/5, and T)6)~=1/20, the termination number 7=8 and sample size n 25 are obtained. Thus the desired noureplacement plan is as follows Start the life test with n=23 units of produet. Do not replace any units that fail. ‘The lot is considered to have met the acceptability criterion if 3 failures have not occurred by 500 hours and the life test is prminated at this time. The lot is considered to have failed to meet the acceptability eriterion if the Sid failure occurs before 600 hours and the teat is terminated at the Gime of the sd failure 26.2 Life Test Plans When Testing With Replacement. From table 2C-4, values of the termination number rand the sample size n may be obtained when testing with replacement for values of a—0.01, 0.05, 0.10, and 0.25; B~0.01, 0.05, 0.10, ancl 0.25; 8,/8)~2/8, 1/2, 1/8, 1/5, and 1/10; and 7/9,— 1/3, 1/5, 1/10, and 1/20. ‘The values 4/4 and 4 shall be computed and for the appropriate values of « and 8, values of r and n shall then be obtained from table 20-4 comple 2-4 Life Test Terminated at Preassigned Time Testing Wath Replacement Example: Find a replacement life test sampling plan which is not to exceed 500 hours and which will accept a lot with mean life of 10,000 hours at feast 90 percent of the time but will reject a lot with mean life of 2,060 hours about 90 percent of the time. bution: In this problem, 7=500, 6= 10,000, 8,~2,000, @=0.10, and 8=0.10. Hence, 6/8 =1/5, '@,~1/20, the termination mumber r=3, and the sample size n=22 are obtained. ‘Thus the desired test plan is.as follows: Start the life test with » ~ 22 units of product, As soon as a umt of product fails, replace it with a new unit, The lot is considered to have met the acceptability criterion if 3 failures have not occurred by: 800 hours and the life test is terminated at this time, ‘Phe lot is considered to have failed to mect the acceptability criterion if the 34 failure occurs before 500 hours and the test is ter- tminated at the time of the 3d failure 2.50LIFE TESTS TERMINATED AT PREASSIGNED TIME. ‘TESTING WITHOUT REPLACEMENT Table 8-3 ‘afe Teal Sampling Plans for Specified o, 8, 6lt, and Tid, mm De 1 mom |e |e mm 2-005 213 uur 27s | 95] 20 447 8431639 ae ao 657 | 33} 90138288409 13 13216 |) 13] 30 4583180 ys | ai eae) ino 6 2] 4] 6 9 1 2% = 0.05, a= 005 p= 005 28 sz 1632 | 67 | 198 90557516 wa ms 472, | 23] 5090108328 18 so? |] wo] 2 ake 4s 3 | os] 8 oom at 10 0 8) 8) ew 0.10 2-005 0.10 ays || | 234 3506751307 | 55] 150 245 40205 2 || 30} 72 109 202 a0 || 19] 47 72g 13 3] 2 37 678 || egw us i oe feo al tao! ig: ol GO ge ae ss se e~0.01 9=025 2=005 s 2 || oo] 12 28 465 a0 |} a5 | 98 a7 768 iz | ml a om are | a] aos te ys | wl 1 2 4 7 | 6) tes us sj 6 9 1 wi] s| 4 5 9 WF mo} 3s] a 4 6 wi 2) 2 2 «4 8Table £C-3—Continued Ti. 90.25 B00 an a8 a6 6991308 ss ia |} a0 "07 280 te ju 75 MS 10s ts se the | 2» ‘iil 8=0.05 og | 2] ws 2 ae sm ia || is] ‘a 73 iar 3a " a | ‘al Fa ‘ 36 20 iH 20 = efale oem miele es mm Mee eee es oe 252— us, ano 213 1/2 1a 1s. ato LIFE TESTS TERMINATED AT PREASSIGNED TIME TESTING WITH REPLACEME! Table 20-4 Life Test Sampling Plane for Specified a, 8 \T nly, and 7 ES m0 on 7051501 m1 488 76153 3 65, 1 a sar 1082 ist a4 Be 108 1939 8 6 10 433887 1448. 399 ao 8 6 o=0.25 28 SIT 76183. 2% 8 eG a7Table #C-4—Continued 0.01 a 197 320 659 tsi |] 52] 140 3440099 26| 50 ws ter ve |] az] az 70a a at) aes an | ae Sh anteioo sO len a a Mle ei se 3) 3 5 2 | 2] 2 4 9 . 5005, o=0.25, 05 2g || 52] 128 14 429 aso || az] at 40280580 1 w3| 58 ot 128 BG | tk) B48 T us a] 3 3 46 5] 10 16 33 aT us 4] 5 8 Wom ist sti nie tenon Ig Ot s eenseeaIG iaa aeeeee a aee a=0.10 s=0.10 27025 s=0.10 28 41] 99 165 330 aso} 3] 58 9819602 102 205, 8) 7 2% 599 83 fu i es a Fd Oe ie 10 yoda 8 o=0.25 20.25 025 376 % 705100 108 ROH ie iss a? 34 atone 2 tat eae 10 Cas No operating cbarntrtle curves re provi for thee oping pas. owevr, wo pols the OC carves LIFE TESTS TERMINATED AT PREASSIGNED TIME—TESTING WITHOUT REPLACEMENT Part 111 LIFE TEST SAMPLING PLANS BASED ON FAILURE RATES 2C7_ FRACTION OF LOT FAILING BEFORE SPECIFIED TIME ‘The sampling plans in this part of the handbook may be used when either (1) the proportion of lot failing before specified time, or (2) the failure rate during time period is specified. Since the sampling plans are based on proportion of lot failing, when the failure rate for period of time is specified, the failure rate shall be multiplied by the specified length of time. ‘That is, poor where p is the proportion of lot failing before specified time T and @ is the failure rate during period of time T.2C8. LIFE TEST SAMPLING PLAN A lot may be considered satisfactory if the feilure rate is less than or equal to py and it may’ be considered unsatisfactory if the failure rate is greater than or equal to p,, where p;>>py. From table 2U-5, values of the termination number r and the factor D are obtained when testing withwut replacement for values of a=0.01, 0.05, and 0.10; 6=0.01, 0.05, and 0.10; and pi/pe=155, 2, 2.5, 3,4, , and 10. Tn this table the sample size, n, is given by [D/p] which means the greatest integer less than or equal to Dips, e.g., [3.8]=3 and [4]=4 Example 20-5 Life Test Terminated at Pre gned Time Testing Without Replacement Example: Find a nonreplacement life test sampling plan which will accept at least 90 percent of thelots for which the failure rate for period of time, expressed as a percentage, is less than or equal to I percent per 1,000 hours and will reject at least 95 percent of the lots for which the failure rate is greater than or equal to 10 percent per 1,000 hours. Solution: In this problem, G.=0.01/1,000=0.00001, 00010, @=0.10, 8=0.05, and T=1,000 hours. Thus, ps=1,000 G.= 0.01, p=1,000 @.=0.10, and pr/py=10. "Looking in table 2C-5, it is seen that r=2 and n=[0.532).01]=53. Thus, the desired plan is as follows: Place 53 units of product ‘on test. If 2 failures occur before time T,, reject the lot and terminate the test at the time of occurrence of the second failure. If one or fewer failures have occurred at time 7, accept the lot and terminate the tent, 0.10/1,000: LIFE TESTS TERMINATED AT PREASSIGNED TIME ‘TESTING WITHOUT REPLACEMENT Table £¢-5 Life Test Sampling Plans for Specified a, 8, and pulpy Values of r (upper numbers) and of D (lower numbers)* 95 o n 82 | “0 me | Sea so | fo | Bo 33 2 26 18 16 waz | is w7 | ae 10.3 10 1“ 18 a ° ne} 846 tos | 702 543 3 10 n 8 6 200) 549 ro2| 460 3.18 ° 7 7 5 4 470] 32] 26 390] 243 17 7 5 4 8 4 3 32| Lor] 137 243) 1.75 110 4 3 3 3 2 2 L37 818 ee saz iSECTION 2D SEQUENTIAL LIFE TEST SAMPLING PLANS 2D1 SEQUENTIAL LIFE TEST SAMPLING PLANS This part of the handbook describes the procedures for use in determining lot acceptability with sequential life tests, Two procedures are given: (1) procedure when testing without replacement, and (2) another procedure when testing with replacement 2D1.1 Use of Sequential Life Test Plans. To determine whether the lot meets the acceptability eti- terion with respect to average length of life, the applicable sampling plan shall be used in accordance with the provisions of section 24, General Description of Life Test Plang, and those in this part of the handbook 2D1.2 Drawing of Samples. All samples shall be drawn in accordance with paragraph 1A0 2D1.3 Sample Plan Code Designation. ‘Ihe sample plan code designation shall be selected from table 2A-1 2D2_ SELECTING SEQUENTIAL LIFE TEST PLANS FOR DETERMINING LOT ACCEPT.- BILITY 202.1. Master Sampling Table. ‘The master sampling table for sequential life test plans for determin= ing acceptability of x lot is table 2D-1 202.2 Obtaining the Sampling Plan. ‘The sequential life test sampling plan consists of a sample size n, the acceptance line intercept hy, the rejection line intercept A, and the common slope « of the two lines. ‘The sampling plan is obtained from the master table 2D2.2.1 Sample Size. ‘The minimum number of units of product that shall be placed on test, re, when testing without replacement, is shown in the master table. When testing with replacement, the sample may be of any size. Increasing the sample size, in either sampling with or without replaceinent, will decrease the time required 10 reach a decision ws to lot acceptability 2D2.2.2 Acceptance Line Intercept. ‘The acceptance line intercept hy is obtained from the master table by multiplying the entry corresponding to the sample plan code designation by the acceptable meat lile 6, 2D2.2.3 Rejection Line Intercept. ‘The rejection line intercept fy 18 obtained from the master table by multiplying Ue entey corresponding to the sample plan code designation by @. 2D2.2.4 Slope of Decision Lines. ‘The common slope x of the acceptance and rejection lines is obtained from the master table by multiplying the entry corresponding to thesample plan code designation by dp, 2D2.3 Acceptance Time. ‘The acceptance time ot ke where ho is the acceptance line intercept obtained in paragraph 2D2.2.2 is the slope of the decision lines as obtained in paragraph 2D2.2.4 A ia the number of failed units of product observed in the length of time that the life teat hae been in progress shall be computed for k=0, 1, 2,2D2.4 Rejection Time. The rejection tin Aybks where A, is the rejection line intercept obtained in paragraph 2D2.2.3 ‘sis the slope of the decision lines as obtained in paragraph 2D2.2.4 & is the number of failed units of product observed in the length of time that the life test has been in progress shall be computed for k=0, 1, 2, Negative values of the rejection time mean that more failures must occur before the rejection of the lot is allowed and hence may be disregarded. 2D3. LOT ACCEPTABILITY PROCEDURES WHEN TESTING WITHOUT REPLACEMENT 2D3.1 Total Life. The acceptability of a lot with respect to a sequential life test shall be judged by the quantity #(0). 2D3.2 Computation. The following quantity shall be computed from the test results: Feat (Ot where 2. denotes the time of the i’th failure in a sample of size n t denotes the length of time that the life test has been in progress k denotes the number of failed units of product observed in time t V(t) denotes the total length of time survived by all units of product on test, failed and unfailed, up to time 2D3.3 Acceptability Criterion. Compare the quantity V(t) with the acceptance time, mentioned in paragraph 2D2.3, and with the rejection time, mentioned in paragraph 2D2.4. If V(t) is equal to or greater than hy-+ks, the lot meets the acceptability criterion; if V(0) is less than or equal to h,+ks, the lot does not meet the acceptability criterion; and if hi +ks0 but the number of failures must exceed ~h,/s. If decisions as to lot acceptability can be made continuously in time, the ereatest 7 287wvings in expected waiting time and the number of failures required for decision are realized over the procedures of Sections 2B and 2C. However, if acceptance of the lot has not been allowed at time ha/n, computation of V(, mentioned in paragraphs 2D3.2 or 2D4.2, may be made periodically. ‘The com: putation of V() must follow each failure but may be made oftener in order to reduce the waiting time required for decision. 2D6 TRUNCATION OF SEQUENTIAL LIFE TESTS ‘The sequential life test, when testing without replacement, will terminate, at the latest, when all units of produet placed on test have failed; the sequential life test, when testing with replacement, should not be allowed to run indefinitely but may be terminated by the procedures of this paragraph Incither case, the sequential life test may be terminated and the lot considered to have met the acceptability eriterion if the number of failures is less than ro, where the value of ry is obtained from table 2D-1, and if— VO zminlhe-+ks, sre) where V(0 and s are explained in paragraphs 22.2.4 and 2D3.2 or 2D4.2 and min (hy +ks, srs) means the lesser of either /ta-+ks or srs. ‘The sequential life test shall be terminated and the lot considered to have failed to meet the acceptability ‘criterion if Vi) hi +-ke or if the number of failures equals ry and Vit)nt>870.75+612.9 k holds and ia stopped ae soon as the inequality does not hold. If 20 unite of product are placed on teat, the inequality may he writtan— —55.90-+30.04 k>1>43.54-+30.04 k If, at the time of stopping, t is less than the left-hand member of the inequality, the lot is considered to have failed to meet the acceptability criterion; if, at the time of stopping, tis greater than the right-hand member of the inequality, the lot is considered to have met the acceptability criterion. Ezample £D-# Expected Number of Failures and Waiting "ime Example: Determine the expected number of failures and the expected waiting time required for a decision in the sequential life test plan of example 2D-1 if n=20 and the mean life of the lot is 0, @,, 8, ‘and @ hours. Solution: From table 2D-1, for sequential plan B-4, Ex(r)—1.8 unite of product Ex(r) =3.0 units of product Ez '=2.6 units of product ) =0.9 unite of product, ‘and from the formulas of paragraph 2D9, E,(t)=0 hours Ex()=310(3.0)/20=48.5 hoursBzample $D-3 Sequential Life Teat Plan Ezample: In the problem of example 2D-1, suppose that « sample of size 20 is placed on test. Units of product that fail are replaced immediately by new unita of product drawn from the same lot. The life test is started at time t=0 and the first five failures occur at 2,,2=25 hours, z3,0=55 hours, 5,970 hours, 24,20=100 hours, and 2, »=160 hours, all times being measured from ¢=0. (a) Verify that no decision has been reached by time 25». (@) Verify that if the sixth failure has not occurred at 196.74 hours, measured from ¢ may be terminated at that time with acceptance of the lot. the life test Solution: The acceptance procedure, as described in section 2D7, data is still within the two decision lines at time 25. acceptance line when k=5 at time t=43.54 +30.64(5 has not yet occurred, the life test may be terminated at t= is drawn in figure 2D-1. The plotted 200, but crosses the V(o/n. Since the sixth failure 196.74 hours with the acceptance of the lot. Remark: If the sixth failure had occurred at t=225 hours, the time saved by making decisions continuously in time is 225—196.74=28.26 hours. Thus, if V(®) is computed only after occurrence of a failure, the life test. would have been prolonged 28.26 hours needlessly. Bzample #D-4 Truncated Sequential Life Test Plan Example: In the problem of example 2D-1, suppose that the sequential life test plan is truncated and a sample of size 20 is placed on test at time f=0. The first 12 failures occur at: 25, 55, 70, 100, 160, 190, 200, 225, 235, 290, 320, and 335 hours. Verify that the lot does not meet the acceptability criterion. Solution: The acceptance procedure, as described in section 2D7, is drawn in figure 2D-2. ‘The plotted data crosses the vertical line below the point (ro, srs) or (12, 7354.8) so the lot does not meet. the nccepta- bility criterion. Remark: While the acceptance of the lot in example 2D-3 is made between failure times 2, x and Zy.x, rejection of the lot in example 2D-4 is made at the failure time zis... This illustrates the point that acceptance of a lot can always be made between failure times, whereas rejection of a lot can only be made at a failure time,vt) Fag, 20-2 WEAMPLE OF SEQUENTIAL LIFE TesT 5000 3000 2000 + qo” 1 2 = 4 cy Nunber of Fatlures (k) 2.61FIG, 20-2 EXAMPLE OF TRUNCATED SEQUENTIAL LIFE TEST 8000 Ee Tegn are) v(t) 4000 oY CT TST TSS TH HOH ReeHite Number of Failures (Ic) 2.62Table #D-1(0) Master Table for Sequential Life Tests ~ e=001 ‘onde a 5 5 Pin Ra a Aq 3 0002 | —o.01g1 | 0.0222 o8 os 03 oo a2 6 908 | —.1777 1292 4 “7 10 1 a 8 2088 | — "4019 2234 18 24 16 3 a 2 aan | —Leatr 2080 24 ay 2a ‘ AS. 15 | 4367 | =. 8571 3491 25 a8 al cd AG 18 5483 0762 3034 27 46 38 9 a a1] eso 2766 1283 a0 a2 45 aa Ae 24 7520 | 1. 4760 3587 32 58 33 La ay a 8479] 1.6643 3863 34 86 60 16 alo su gauy | <1 say 3067 a6 us 68 19 an 4s vasaz| 2 6a 5808 4s 108 10.5 a2 Al? go] 7158) —3 3670 6350 53 138 3 46 a3 75| 2.0330) | —3 9004 6808 60 m0 Ig eo at 9} 2.3305] —4 572 oe 66 20.2 20 14 As v0] 2.8477 | | 8 5806 7336 76 24 m6 104 a6 wso| 3.3127 | 6 S022 7800 a6 327 37.3 Ine AAT zs} 43142| — 4679 020 106 483 a8 2.1 ae 300 | 5.408 | —10. 1100 sare ine 03.5 Tar ae Table #D-1(8) Master Table for Sequential Life Teste 90.05, Sd Code " 4 ue a Fan) a m0 ra B 3| cos] —oo6s0| a. 08s9 os os o4 B2 6 2254 | —, 2804 2400 Le 16 Lt Ra a doug | = 5280 3405, 18 23 19 Ba 2 805 |=. 7453 4086 18 30 26 BS 1% 74s | —! 9430 4576 21 a7 33 Do 16 642 1952 4972 23 a3 41 BT a 1.0209] —1 3107 15282 25 50 43 BB 2) a495 | 1.4757 3538 27 56 55 BY al xany| Shes 9798, 28 6s Bs B40 30) aoe | 1.7266 5048 30 60 10 Bul 45] oir | 2 4523 8607 37 10.0 107 BA? 60} —2.3620| —3. 0325 024 43 1B 45 Bug 75| 2716) —3 5327 707 43 161 182 Bu | a127| | —4 0079 7530 5B 19.2 m1 Bus 120) 3.7522] —4 8173 733 62 260 25 BAG 150} 4asi4| | —5 5010 8053 69 31.0 ar. Bar z25| s.sa80| —7. 1100. sao as ao sa Bas 00] 6.8773) —R 4aee 8600 a8 oat 751Table #0-1(¢) Master Table for Sequential Life Teats =0.10 co c2 3 co cs C8 cr cs co C10 cu 12 1 14 C15 C16 car cB Por epson of he ode, se pr 2432 3 6 9 ra 15 18 2 Pa 7 30 rr oo % %0 120 150 25 300 or at 14 16 18 20 22 24 25 a7 33 7387 38 7642 a3 7827 47 8105, Ba 201 60 8588 a4 8i74 ac Table #D-1(@) 0.25, eoneenenere SBE Sueeeeeyene geevers Table for Sequential Life Tests peppees Been 185 m4 38.0 ar BEESBSESaoeee ewe 08 16 21 a7 aa 39 a5 a1 56 62 80 1.8 05 7.2 26 m0 aud 541 Eun ° Been 35 40 46 52 a1 ma 40 10 2.0 43 50.5, "Yor exlaaton te code, separ, 243.2408 table 24-1APPENDIX 2D-1 In figure 2D-3, a comparison is shown in the expected number of failures required for @ decision between a life test terminated at preassigned time when sampling with replacement plan of section 2C and that for a sequential plan of section 2D. Both plans have OC curves passing through the points (6=1500, 1-a=0.95) and (6,=300, B=0.10) In figure 2D-4, « comparison is shown in the expected waiting times required for a decision between Alife test terminated at preassigned time when sampling with replacement plan and a sequential replacement plan. Both plans have sample size n=-20 and have OC curves passing through the points (8.=1500, 1 a=0.95) and (#,—300, p=0.10). ‘The above comparisons are typical of the savings in time and number of failures required for a decision.‘otto = d pue *S0°0 =p ‘00€ = Te ‘oost = °@ aeuy Yons ©, UBT YoRe ZOZ eA gp oUy (@) GAIT KVat 0g 000 o0gz cooz cost 000T cos ° T a 1. r | 2 eee = — ——_ . - t | Weld TeFquenbeg | _ : __| : 5 ae | 4 | : |. | a | | | L ¢ Wels qusnsoe ey ua TUTE ‘oupy, Peustsse-ery ye pozeUTULE] 380], GFT] i a SNVTd INSHHOVTdHY HLIM ONITAKVS “SMILE OSNOISSV-Mid JV CAIVNIMUSL ISH, 417 QHY TVLINENDSS Wl NOISICUC Wa CHUTNEY SHUNTIVA AO UBAHON CELOEGKR JO NOSTUVaHOO eae “olasot'o =d Bue *S0°0 =” ‘0OE ='9 ‘00ST = °@ 7eUa Yone ST UETA yore HOF eAaINo Op eu (8) SIT NVaH o0se 0006 ose ooze 00ST 000T 00S ° UITA UPL Tey UeNbeg t———— UeTg quoueoeTdey yztM BuyTdueg ‘omy, poudtsse-oig FE popeUTLIO] 7605 = — ot | | I —— ott SNVId INGHIOYIGTY HLIM ONTTHHVS ‘SLL GANDISSV~dud LY CALYNIWERL Isal sat ONY ‘TVLINE(DSS Od NOISIOAC wd CSULNCEY ALL ONTLIVM CAL AACXa JO NOSTUVdHOD ae -

108 Sampling Procedures

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

108 Sampling Procedures

Uploaded by

Copyright:

Available Formats

You might also like