《BS-IEC-60605-6-1997.pdf》由会员分享，可在线阅读，更多相关《BS-IEC-60605-6-1997.pdf（14页珍藏版）》请在三一文库上搜索。
1、BRITISH STANDARD BS IEC 60605-6:1997 Incorporating Corrigendum No. 1 Equipment reliability testing Part 6: Tests for the validity of the constant failure rate or constant failure intensity assumptions ICS 29.020 NO COPYING WITHOUT BSI PERMISSION EXCEPT AS PERMITTED BY COPYRIGHT LAW Licensed Copy: sh
2、effieldun sheffieldun, na, Sun Nov 26 11:50:12 GMT+00:00 2006, Uncontrolled Copy, (c) BSI BS IEC 60605-6:1997 This British Standard, having been prepared under the direction of the Management Systems Sector Board, was published under the authority of the Standards Board and comes into effect on 15 S
3、eptember 1997 BSI 07-2001 ISBN 0 580 28095 0 National foreword This British Standard reproduces verbatim IEC 60605-6:1997 including Corrigendum December 2000 and implements it as the UK national standard. The UK participation in its preparation was entrusted by Technical Committee DS/1, Dependabilit
4、y and terotechnology, to Subcommittee DS/1/1, Dependability, which has the responsibility to: aid enquirers to understand the text; present to the responsible international/European committee any enquiries on the interpretation, or proposals for change, and keep the UK interests informed; monitor re
5、lated international and European developments and promulgate them in the UK. A list of organizations represented on this subcommittee can be obtained on request to its secretary. From 1 January 1997, all IEC publications have the number 60 000 added to the old number. For instance, IEC 27-1 has been
6、 renumbered as IEC 60027-1. For a period of time during the change over from one numbering system to the other, publications may contain identifiers from both systems. Cross-references The British Standards which implement international or European publications referred to in this document may be fo
7、und in the BSI Standards Catalogue under the section entitled “International Standards Correspondence Index”, or using the “Find” facility of the BSI Standards Electronic Catalogue. A British Standard does not purport to include all the necessary provisions of a contract, User of British Standards a
8、re responsible for their correct application. Compliance with a British Standard does not of itself confer immunity from legal obligations. Summary of pages This document comprises a front cover, an inside front cover, pages i and ii, the CEI IEC title page, page ii, pages 1 to 7 and a back cover. T
9、he BSI copyright notice displayed in this document indicates when the document was last issued. Amendments issued since publication Amd. No.DateComments 13096 Corr No. 1 July 2001Indicated by a sideline Licensed Copy: sheffieldun sheffieldun, na, Sun Nov 26 11:50:12 GMT+00:00 2006, Uncontrolled Copy
10、, (c) BSI BS IEC 60605-6:1997 BSI 07:2001 i Contents Page National forewordInside front cover Text of CEI IEC 60605-61 Licensed Copy: sheffieldun sheffieldun, na, Sun Nov 26 11:50:12 GMT+00:00 2006, Uncontrolled Copy, (c) BSI ii blank Licensed Copy: sheffieldun sheffieldun, na, Sun Nov 26 11:50:12 G
11、MT+00:00 2006, Uncontrolled Copy, (c) BSI Licensed Copy: sheffieldun sheffieldun, na, Sun Nov 26 11:50:12 GMT+00:00 2006, Uncontrolled Copy, (c) BSI BS IEC 60605-6:1997 ii BSI 07:2001 Contents Page Introduction1 1Scope1 2Normative references1 3Definitions1 4Symbols1 5Requirements2 6Tests for constan
12、t failure rate2 7Test for constant failure intensity3 Annex A (normative) Chi-square distribution5 Annex B (informative) Examples: Test for constant failure rate6 Annex C (informative) Example: Test for constant failure intensity7 Annex D (informative) Bibliography7 Figure B.1 Auxiliary function F(i
13、,11) as a function of time7 Table 1 Critical values of U as a function of ? ?4 Table A.1 ?2 distribution5 Table B.1 Twenty ordered times to failure out of 40 tested items6 Table B.2 Ordered times to failure and computed auxiliary function6 Table C.1 Eight times at which item failures occurred7 Licen
14、sed Copy: sheffieldun sheffieldun, na, Sun Nov 26 11:50:12 GMT+00:00 2006, Uncontrolled Copy, (c) BSI BS IEC 60605-6:1997 BSI 07:2001 1 Introduction The techniques given in this part of IEC 60605 for testing constant failure rate or constant failure intensity assumptions are numerical and graphical
15、procedures that can best be implemented through the use of a computer. More generally, these techniques can be used for testing the assumption that events are exponentially distributed. 1 Scope This part of IEC 60605 specifies procedures to verify the assumption of a constant failure rate or constan
16、t failure intensity as defined in IEC 60050(191). These procedures are applicable whenever it is necessary to verify these assumptions. This may be due to a requirement or for the purpose of assessing the behaviour in time of the failure rate or the failure intensity. The tests specified in this Int
17、ernational Standard are one of the following: to test whether the times to failure of items are exponentially distributed, i.e. the failure rate is constant; to test whether the times between failures of a single repaired item do not have any time trend, i.e. the failure intensity does not exhibit a
18、n increasing or decreasing trend. 2 Normative references The following normative documents contain provision which, through reference in this text, constitute provisions of this part of IEC 60605. At the time of publication, the editions indicated were valid. All normative documents are subject to r
19、evision, and parties to agreement based on this part of IEC 60605 are encouraged to investigate the possibility of applying the most recent editions of the normative documents indicated below. Members of IEC and ISO maintain registers of currently valid International Standards. IEC 60050(191):1990,
20、International Electrotechnical Vocabulary (IEV) Chapter 191: Dependability and quality of service. IEC 60300-3-4:1996, Dependability management Part 3: Application guide Section 4: Guide to the specification of dependability requirements. IEC 61014:1989, Programmes for reliability growth. IEC 61164:
21、1995, Reliability growth Statistical test and estimation methods. IEC 61649, Procedures for goodness-of-fit tests, confidence intervals and lower confidence limits for Weibull distributed data. ISO 3534-1:1993, Statistics Vocabulary and symbols Part 1: Probability and general statistical terms. 3 De
22、finitions For the purpose of this part of IEC 60605, terms and definitions are in accordance with IEC 60050(191). 4 Symbols F(i,n)auxiliary function for the graphical procedure when testing n items for constant failure rate nsample size, the total number of items being tested for constant failure ra
23、te tithe time corresponding to the i-th ordered failure, used when testing n items for constant failure rate dparameter related to number of relevant failures; if the validity test is done at a point in time coinciding with a failure, then d = r 1; if not then d = r mnumber of intervals when using t
24、he large sample test wwidth of the interval measured in accumulated time ?risk of wrongly rejecting the assumption that the (instantaneous) failure rate or the (instantaneous) failure intensity are constant, when they really are constant rthe number of relevant failures during the test T*total test
25、time accumulated Licensed Copy: sheffieldun sheffieldun, na, Sun Nov 26 11:50:12 GMT+00:00 2006, Uncontrolled Copy, (c) BSI BS IEC 60605-6:1997 2 BSI 07:2001 5 Requirements In order for the procedures specified in this standard to be valid, it is necessary that the following requirements be satisfie
26、d: When testing n non-repaired items for the constant failure rate assumption: for the numerical procedures, at least 10 times to failure are available; for the graphical procedure, not less than four times to failure are available. When testing a single repaired item for the constant failure intens
27、ity assumption: the item shall be tested for a sufficiently long time so that at least six times between failures are available. NOTE 1General guidance for these procedures is given in IEC 60300-3-4 and future IEC 60300-3-5 (see Bibliography). NOTE 2In this standard, the term “time” can refer to len
28、gth, cycles or other quantities. The term “failure” can also refer to other specified events such as repair completion or any other particular event. 6 Tests for constant failure rate This clause applies when a sample of n items is put on test. For large samples, numerical procedures are given in 6.
29、1 and 6.2. For very small samples, only a subjective graphical method is available, given in 6.3. The operating environment shall be the same for all the items tested. At the end of the testing period, not all the items will have necessarily failed. There will be a total of r recorded relevant times
30、 to failure. Order the times to failure in increasing order of magnitude, and denote the ordered sample t1, t2,.,tr. For i = 1 to r, compute the accumulated time to the i-th failure as: Compute the accumulated total test time as: 6.1 Sample size between 10 and 40 (numerical procedure) Compute the te
31、st statistic: Compare this quantity ?2 to the theoretical values of?2(v), tabulated in Table A.1, using v = 2d. Perform a two-sided test, for a 10 % significance level, using Table A.1, as follows: If then reject the assumption of a constant failure rate. The failure rate is likely to be increasing.
32、 If then reject the assumption of a constant failure rate. The failure rate is likely to be decreasing. Otherwise, if neither of these inequalities holds, do not reject the assumption of a constant failure rate. Tiaccumulated time of the i-th relevant failure Traccumulated time of the last failure u
33、pthe p fractile of the cumulative standard normal distribution Ucalculated value of the statistic, used when testing for constant failure intensity Eexpected number of failures in a time interval Oiobserved number of failures in the i-th time interval vnumber of degrees of freedom Ti tkni?ti+ k1= i
34、? = T* tknr?t+ r k1= r ? = Licensed Copy: sheffieldun sheffieldun, na, Sun Nov 26 11:50:12 GMT+00:00 2006, Uncontrolled Copy, (c) BSI BS IEC 60605-6:1997 BSI 07:2001 3 6.2 Sample size larger than 40 (numerical procedure) Divide the time between zero and the total accumulated test time T* into m equa
35、l intervals of width w. The expected number of failures in each interval is: m shall be chosen so that E is equal to or greater than 5. Compute the test statistic: Compare the calculated value of ?2 to the theoretical value of ?2(v) listed in Table A.1, using v = m 1. Perform a one-sided test, for a
36、 10 % significance level, using Table A.1, as follows: Ifthen reject the assumption of a constant failure rate. It is not possible to assess whether the failure rate is increasing or decreasing with this procedure. Otherwise, do not reject the assumption of a constant failure rate. 6.3 Sample size b
37、etween 4 and 9 (graphical procedure) The graphical procedure is subjective. The risk (?) of wrongly rejecting the assumption that the failure rate is constant therefore cannot be quantified. Plot the sequence of ordered times to failure t1, t2,., tr along the linear scale of a semi-log paper. The lo
38、garithmic scale is used for the auxiliary function F(i,n), where i is the index of the corresponding failure time ti and n is the sample size: If the plot of this function looks linear on the semi-log paper, then there is no evidence to reject the assumption that the failure rate is constant. If it
39、does not look linear then this assumption should be rejected. 6.4 Action to be taken if the constant failure rate assumption is rejected If the constant failure rate assumption is rejected, it is recommended to analyze the data further to determine the possible cause. Numerical analysis should, wher
40、ever possible, be supported by physical investigations and engineering considerations. The items may be subject to wear-out in the time interval considered or a mechanism inducing early failures may be present. There is also the possibility that the items do not come from a homogeneous population, i
41、n which case there may be a mixture of several (constant) failure rates. All these situations deserve further investigation. If wear-out or early failures are suspected, the procedures of IEC 61649 should be used. If, on the other hand, a mixture of populations is suspected, efforts should be made t
42、o identify and separate the different populations, and to analyze these separately. Whatever the cause for the rejection of the constant failure rate assumption, compliance methods that require this assumption should not be applied. NOTEIf the constant failure rate assumption is not rejected, the co
43、nclusion is that the times to failure have not been proven to deviate from the exponential assumption. 7 Test for constant failure intensity This clause applies when a single repaired item is tested. Testing for constant failure intensity implies that times between successive relevant failures exhib
44、it neither an increasing nor a decreasing trend. If so, the item can be considered as being renewed after each repair. In this case, and only in this case, these times between failures can be considered as times to failure. IEC 1014 gives information concerning relevant failures and other related de
45、finitions. Ew d T * - -= F i n,? n0,4+ ni0,7+ - -= Licensed Copy: sheffieldun sheffieldun, na, Sun Nov 26 11:50:12 GMT+00:00 2006, Uncontrolled Copy, (c) BSI BS IEC 60605-6:1997 4 BSI 07:2001 7.1 Number of failure times is 6 or larger (numerical procedure) This numerical procedure requires that ther
46、e are at least six relevant successive failures recorded during the testing time T*. The accumulated time to the i-th failure is Ti. The times between failures are Ti + 1 Ti, i = 1 to r. This procedure can be applied either at the time of the last failure Tr or at any other later time T* during whic
47、h the item continues to perform its function. Step 1 For each relevant accumulated failure time Ti compute the quantity U: If T* Tr then: If T* = Tr then: Step 2 Specify the acceptable risk ? to wrongly reject the assumption of constant failure intensity, given that it really is constant. Recommende
48、d values of ? are given in Table 1. Table 1 Critical values of U as a function of ? Step 3 Reject the assumption of constant failure intensity if absolute value of U is greater than the, critical value given in Table 1. Otherwise, the assumption is not rejected. Under the no-trend assumption, i.e. a
49、ssuming the failure intensity is constant over time, the U statistic follows the standard normal distribution. Large absolute values of U constitute evidence to reject this assumption. Large positive values of U occur whenever there is a decreasing trend in times between successive failures. Conversely, large negative values of U occur whenever these times have an increasing trend, i.e. they become longer since the failure intensity becomes smaller. NOTENot rejecting the constant failure intensit