1、 i 2 I EIA RB4-A 69 W 3234600 O006329 O W / RELIABILITY BULLETIN No.4-A Reliability Quantif ication . MAY 1969 PREPARED BY THE 6-41 COMMITTEE ON RELIABILITY Engineering Departmen t. ELECT.RONIC . INDUSTRIES ASSOCIATION I. 2001 EYE STREET, N. W. WASHINGTON, D. C. 20006 COPYRIGHT Electronic Industries
2、 AllianceLicensed by Information Handling ServicesEIA RB4-A 69 = 3234600 O006330 7 I,. i 5- RELIABILITY QUANTIFICATION BULLETIN A GUIDE FOR THE USE BY COMPANIES CONTRACTING FOR DESIGN OF EJXCTRONIC PRODUCTS WITH THE: DEPARTMENT OF DEFENSE (DOD) AND OTHER GOVERNMENT AGENCIES Prepared by the Reliabili
3、ty Quantification G-41 Committee on Reliability Task Group May 1969 Electronic Industries Association Washington, D. C. 20006 Engineering Department 2001 Eye Street, N.W. COPYRIGHT Electronic Industries AllianceLicensed by Information Handling ServicesEIA RB4-A 69 m 3234600 0006331 9 m FOREWORD This
4、 Bulletin presents concepts and techniques for quanti- fying electronic equipment reliability. The techniques are responsive to the requirements of various branches of the Department of Defense and are also useful with regard to other Government agencies (e.g., NASA). This Bulletin has been publishe
5、d by the Electronic Industries Association, in the belief that the approach, as defined, will interest and benefit those responsible for reliability quantification programs. This Bulletin was prepared by a Special Task Group (68-002) of the Electronic Industries Assoeiation G-hl Reliability Committe
6、e consisting of: G. L. Hetzel Bell Telephone Labor at or ies C. K, Smith Sperry Gyroscope Company D. I. Troxel Radio Corporation of America W. T, Weir - Task Group (68-002) Chairman General Electric Company E. D. Karmiol, Chairman G-41 Reliability Committee COPYRIGHT Electronic Industries AllianceLi
7、censed by Information Handling ServicesEIA RBq-A 69 m 3234600 0006332 O m TABLE OF CONTENTS SECTION Page I INTRODUCTION . 1-1 II III IV v VI VI1 ANALYTICAL ACTIVITIES 11.1 . A . Mission Definition and Analysis 11-1 B . System Analysis . 11-1 C . Integrated Test Plan 11-2 D . Failure Analysis and Cla
8、ssification . 11-2 E . Statistical Model . 11-5 F . Reliability Reporting 11-4 REQUIREMENTS 111-1 ALLOCATION . 1v-1 A . General . 1v-1 . B . Contractual Aspects 1v-2 C . Factors Involved . 1v-2 D . Methods 1v-3 E . Usefulness of Results 1v-4 RELIABILITY PREDICTION . V-1 A . Prediction Concept . V-1
9、B Prediction Factors . V-1 D . Prediction Techniques - General . V-3 E . Prediction Technqiues - Specific V-5 F . Prediction Data Sources V-5 C. Prediction Applications ; V-2 MEASUREMENT v1-1 A- fijodelRequirements v1-2 B . Basic Measurement Model . v1-3 1 . Operating Assumption8 . v1-4 2 . Mathemat
10、ical Model . v1-5 DEMONSTRATION v11-1 A . Introduction vn-1 B . Background and Problems on Standardized Demonstration Methods VII-I iii COPYRIGHT Electronic Industries AllianceLicensed by Information Handling ServicesI- EIA RB4-A b9 - 3234b00 0006333 2 - TABLE OF CONTENTS (Contd) SECTION Page C. Dem
11、onstration Methods VII-2 1. Demonstration by Analysis and Prediction . . . . VII-2 2. Demonstration by Integrated Reliability . . . . . , . . . . . . . . . . . . . . . . VII-3 3. Demonstration by Specific Test . . . a . . . . . . . VII-3 4. , , . . . . . , . . . . . . . . . *. . . . . VII-3 Penaltie
12、s and Incentives Associated with Reliability Demonstration . . . . . . . . . . . . . . . . . . . . . . . . . VII-5 Summary of Reliability Demonstration Quantification Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . VII-6 . . . . . . . . . . . a . . . . . . . Measurements On Site, or O
13、perational Evaluation Data Assessment D. E. VI11 REFERENCES . . . . . . . . . . . . . . . . . . . . . *. . . . . . . b VIII-1 rx APPENDIX . . . . , . . . , . . . . . . . . . . . . . . . . . IX-1 A. IX-1 B. Variable Attribute Error Propagation Method . . . . . M-3 C. Standardization DI-5 Bayesian Mea
14、surement Model . . . . . . , . . , . . . . . . . . . . . . . . . *. . . . . . . . . . . . . COPYRIGHT Electronic Industries AllianceLicensed by Information Handling ServicesU ,. EIA RB4-A b9 m 3234600 0006334 4 m SECTION I INTRODUCTION This reliability Bulletin is provided as a guide for engineering
15、 and management personnel concerned with the quantification of reliability and/or effectiveness. It covers the specification and apportionment of reliability, the essential elements or ana- lytical activities of a quantification program, prediction, measurement, and demonstration. For the purpose of
16、 this Bulletin r diction is based upon data from sources such as MIL-HDBK-21781T, Department of Defense Publi- cation -181080(2), and RADC Reliability Notebook Volume 2(3) e Measurement uses actual test data from component (black box as opposed to piece part, e.g., resistor, capacitor) and higher le
17、vels of assembly to estimate system reliabilit . Measur m nt is described Demonstra- tion may use the same information as asurement or may be predicted on specific tests (e.g., MIL STD 78i($e). penalty for failure to meet acceptance requirements. There may be various states of prediction and measure
18、ment and these can be combined into current dynamic assessments of reliability, using all information, by the Bayesian Measurement Techniques presented in the Appendix of this Bulletin. Reliability and availability are two significant inputs to a system effectiveness calculation. The techniques desc
19、ribed in this Bulletin are intended primarily for use in quantifying reliability in those instances where failure-free operation is required for a defined mission. The techniques are also applicable for determining the reliability inputs required for availability calculation in NAVWEPS OD 29304(4) a
20、nd NAVORD OD 2930 t /Addendum 77 5 . Demonstration implies a e ability is a function of reliability and maintainability Although it is recognized that the human element can be and frequently is a major factor in the reliable functioning of a system, this aspect is not specifically treated in this Bu
21、lletin. Another problem of system reliability which is not specifically treated in this Bulletin is the effect of degradation, in dormant storage, on relia- bility and the surveillance programs required to monitor and control aging. Two techniques for measurement/demonstration are covered in the App
22、endix, They are not included in the main body of the report since there is too little experience with them to recommend them as standard methods at this point in time. They are not omitted altogether since they both offer considerable potential for the future. 1-1 o COPYRIGHT Electronic Industries A
23、llianceLicensed by Information Handling Services” x .% -Q - EIA RB4-A 69 3234b00 000b335 b SECTION II ANALYTICAL ACTIVITIES Quantification of reliability and assurance that design requirements will be met is dependent upon the successful completion of a number of tasks. To varying degrees each of th
24、ese must be accomplished whether the desired out- put is a prediction, measurement, or demonstration. The tasks which should be performed are discussed in A through F: - A, MISSION DEFINITION AND ANALYSIS Reliability is the probability of successfully completing some specified mission. Without a sta
25、ted mission, reliability is an indeterminate quantity, The reliability requirement) provided by the customer include the mission(s) against which system reliability is measured. The apportion- ment provides the reliability allocated to lower level assemblies and the specific mission of each of these
26、 assemblies is defined in mission analysis. Mission parameters such as functional requirements, success criteria, phases of the mission, sequence of events, operating modes, mission environments, and the required mission times are clearly identified, This information is required to determine the tes
27、ting necessary to sim- ulate the mission and to establish the manner in which test information will be used to measure/demonstrate reliability; it also provides a basis for assuring that design information required for total mission definition is considered in test planning. In summary, it is import
28、ant to compare both the design requirement8 and the test requirements with the mission requirements to assure compatibility. . B. SYSTEM ANALYSIS This analysis includes configuration definition, failure modes and effects, preparation of the reliability block diagram(s) , development of reliability c
29、quntion(s) , and reliability allocation. This analysis defines the hardware on which data collection is required. In the case of redundant devices, it is necessary to state whether data are required on each redundant channel or whether system performance is sufEicient. . Success-failure criteria are
30、 established for each hardware element at this state of the analysis. e 11-1 - COPYRIGHT Electronic Industries AllianceLicensed by Information Handling ServicesEIA REY-A 69 3234600 0006336 m C. INTEGRATED TEST PLAN On programs where statements such as “No special testing shall be ac- complished for
31、reliability/demonstrationtf appear in contractual docu- ments such as the Statement of Work, it is necessary to analyze the integrated test plah to dete ine which of the tests being accomplished for other purposes provides 3 ata relevant to reliability measurement. In those cases where demonstration
32、 is required and the integrated test plan is not structured to provide sufficient test data for this demonstra- tion, a special reliability test to provide the additional information re- quired must be undertaken. Among the advantages of analysis of the integrated test plan is the visibility that it
33、 provides. If one component (black boar as opposed to piece part) is receiving several thouisand equiv- alent missions of test and another component only ten this fact is dramat- ically highlighted and reallocation of test reBource8 can be considered without increaeing total coat. On programs where
34、special. reliability tests are required for demonstra- tion (e, g., MIJPSTD-() it is still advantageous to analyze the integrated test plan and to perform the mission and system analyses, The mission and system analyses enable the contractor to determine the relationship between MIGSTD-781 tests and
35、 the actual mission environments. This permits realistic recommendations of the proper test level (Le., A, A-1, B, C, D, E, F, G, H, or J) and a realistic appraisal of the relationship of the test to the mission. Employing the techniques for reliability measurement which are recommended in Section V
36、I of this Standard would provide a timely assessment of the risks of entering the reliability demonstration test. . If the contractor is assigned to develop the reliability demonstration test, a guide manual such as reference 8 should be used in conjunction with the analyses described herein. D. FAI
37、LURE ANALYSIS AND CLASSIFICATION The primary purpose for a failure evaluation and reporting activity is the analysis of failures and elimination of failure causes, However, this activity also provides an important input to a reliability measurement/ demonstration program. It is essential to clearly
38、define failure prior to instituting test, particularly in the case of demonstration programs with incentive aspects. Classification is based upon many factors among which the more signifi- cant are: The success criteria developed during mission analysis, the . TI-2 COPYRIGHT Electronic Industries Al
39、lianceLicensed by Information Handling ServicesEIA RB4-A 69 W 3234600 0006337 T W type and level of the test, the degree and cause of the failure, and the severity of the test relative to the mission. It is important that the ground nilee for elimination of failure be clearly established prior to bi
40、st. Several ground rules are generally applicable: L (1) In any testing, the test time and associated failures are treated collectively. Either both time and failures collected from a test are applicable or neither time nor failures collected from the test are applicable. (This does not mean a failu
41、re cannot be ex- cluded for valid reasons or that test data is not used when failure- free tests are accomplished. ) (2) Failures and test time in the non-mission environment are not counted. (Testing under ambient conditions is normally con- sidered relevant. ) (3) Failures occurring as a result of
42、 mishandling, test error or test equipment error are not normally counted, but the link between reliability and human factors should not be overlooked. E. STATISTICAL MODEL Reliability measurement/demonstration requires a statistical model which permits eetimation of system reliability based upon da
43、ta obtained from in- dependent test of lower assembly levels, as well as system level teat, Many techniques have been developed for this purpose(9). The atandard recornmendod here is that shown in NAVWEPS OD 29304(4), and NAVORD OD 29304/Addendrn(). This method is well documented, flexible, and easi
44、ly understood. The techniques described in NAVWEPS OD 29304 have been applied successfully on Air Force Programs such as Thor, Atlas, Titan, Minuteman, and various Recoverable Satellite Programs. The Navys contractors have applied NAVWEPS OD 29304 on Navy Programs (e. g. , the Strategic System Progr
45、am Office Polaris and Poseidon and. Ship Systems Command Sonar). It has also been applied on programs under the direction of other governmental agencies. In addition to the rather wide and successful use it has enjoyed, on R to indicate status against estimated growth; to pinpoint quantitative relia
46、bility problem areas; and to discuss problem areas indicating corrective action planned and anticipated reliability gained by incorporating corrective action. *The Navy has distributed several thousand copies of OD 29304 and it is cur- rently available from the Superintendent of Documents, U. S. Gov
47、ernment Printing Office, Washington, D. C., 20402. COPYRIGHT Electronic Industries AllianceLicensed by Information Handling Services- .P e I EIA RB4-A 69 3231.1600 0006339 3 = e o The Reliability Status Report is normally issued periodically (volume of data being accumulated determines optimum frequ
48、ency) and contains mission failure rate and reliability results at two confidence levels (usually, but not necessarily, maximum likeli- hood estimate and 80%) for each appr0priat.e hardware level (e. g., component, equipment, subsystem). It could aso report the con- fidence level at which the specif
49、ied reliability has been demon- . strated by each hardware element. The reliability status report contains items such aiS: * A brief description of the subsystem and equipment operation and of the mission for which reliability is reported. The subsystem block diagram and reliability equation, A summary of the test data sources. A table relating the currently measured reliability versus reliability requirements for the Eiubsystern ad equipments. Growth curves for the subsystem and equipments show- ing measured reliability versus time with the required reliabi
