1、E 23261134 0006172 3 m, TIN 45-01 E Page 1 Recommendation T/N 45-01 (Edinburgh 1988) TESTING THE COMPLIANCE OF AN EQUIPMENT WJTH ITS RELIABILITY, MAINTAINABILITY AND AVAILABILITY SPECIFICATIONS Recommendation proposed by Working Group TfWG 14 “Network Aspects” (NA) Text of the Recommendation adopted
2、 by the “Telecomminications” Commission: “The European Conference of Postal and Telecommunications Administrations, considering - that for various equipments, besides the normal functional aspects, availability, reliability and maintainability aspects are specified, - that when accepting these eqiii
3、pments from the manufacturer or deliverer, the Administrations must be sure that these availability, reliability and maintainability specifications are met, - that it is therefore advisable o use testing methods, most of which are based on statistical elaboration of various information in this field
4、; which methods must be agreed with the manufacturer or deliverer of the equipment concerned, - that this subject remains the total responsibility of the Administration concerned, i.e. this Recommendation will not dictate what the Administration should or should not do in this matter, - that it is,
5、however, recommendable to have some guidance on the matter of testing the compliance of the specified reliability, maintainability and availability aspects, O recommends the following guidance, which describes a method for testing the compliance of an equipment with its reliability, maintainability
6、and availability specifications.” Edition of Jiine 30. 1989 , B 2326414 00b173 5 T/N 45-01 E Page 2 1. INTRODUCTION This Recommendation is made of three parts: - the first part indicates the application area of the Recommendation by listing reliability, maintainability, availability characteristics
7、to which the Recommendation applies; - the second part introduces three testing principles which can be used for testing the compliance of an equipment to its reliability, maintainability, or availability specifications and indicates for each principle to which characteristic it is best fitted; - th
8、e third part is devoted to the description of methods and mathematical tools referrenced in the second part. 2. CHARACTERISTICS TO BE VERIFIED The following list gives the main reliability, maintainability and availability characteristics for which a method for compliance testing is proposed in this
9、 Recommendation. 2.1, Reliability characteristics 2.1.1. Global failure rate This parameter is used to evaluate the number of repairs which will have to be done, for a given period, on the considered equipment. 2.1.2. Functional failure rate This parameter is used to evaluate the number of times whe
10、n one equipment is not able to work as specified during a given period. For a switching system, several functional failure rates can be spedified according to the consequence of the considered failure: - failure affecting a given group of subscribers; - failure affecting a given group of circuits; -
11、 failure which do not affect specifically a given group of subscribers or circuits but which lower the trafficability performance of the system. 2.2. Maintainability characteristics 2.2.1. Probability of failed state detection It is the probability that the existence of a failure inside the equipmen
12、t is detected, whether the required functions of the equipment are fulfiled or not. 2.2.2. EfJiciency of localization of a failed item When a failure is detected, a localization procedure takes place which will identify a given set of presumably failed items as a failure localization. The efficiency
13、 of the localization procedure is defined by two probabilities corresponding to the following definition: For precise localization, where no more than k items should be identified per failure (k=l in most applications), a probability P1 is specified. However, as localization procedures cannot be per
14、fect, a larger set (between k+l and, say, m items) can be identified, but with low probalility P2 (consequently, the probability that the number of identified items per failure is more than m is equal to 1 -(Pi +P2). In summary, the efficiency of localization is defined by the two probabilities P1 a
15、nd P2 such that: PI =P (localization among 1 or 2 or . k items) P2 = P (localization among k + 1, or k + 2, . . . or m items) with m h, is less than ). The ratio D = h,/ho is called the discrimination factor. Standard test plans When the values of u, , ho and D are given, one can derive the operatin
16、g test time (T) which has to be accumulated by the equipments or systems and the maximum number (C) of failures occurring during the accumulated test time T compatible with the decision that the equipment complies with its failure rate specification. Corresponding test plans are described in IEC 605
17、-7 Publication. Edition of June 30, 1989 7 r . , i I 2326434 0006376 O T/N 45-01 E Page 5 4.1.3.2. 4.2, 4.2.1. 4.2.2. 4.2.3. 4.2.4. 4.2.5. 4.2.5.1. 4.2.5.2. Other test plans It can be convenient to choose beforehand the value of the accumulated test time T (which allows to decide beforehand of the q
18、uantity of equipments or systems to be monitored and of the duration of the test). In this case, u, , ho and T are given and the value of the discrimination factor ) as well as the maximum number (C) of failures which can occur during the accumulated test time T are derived. The method for deriving
19、D and C from the values of u, , b) reference to quality/reliability programme of system or equipment and components; c) component screening. Any assumption regarding reliability or maintainability growth shall be stated. Environmental conditions. The environmental conditions for which the prediction
20、 is performed shall be those specified for the equipment/system operation. Operational conditions. The operational conditions for which the prediction is performed shall be those stated for the equipment/system in its relevant specification. Definition of maintenance actions. The equipment/system sp
21、ecification defines as maintenance requirements on which equipment/system complexity level corrective maintenance is to be performed, siich as failure localization of replaceable units or failure localization down to component level. Accordingly, the expected mean values of the durations .of the cor
22、responding maintenance actions, when used for the prediction, shall be stated (see par.agraph “Maintainability data”). Edition of June 30, 1989 - - = 232b434 0006377 2 T/N 45-01 E Page 6 4.2.5,3. 4.2.5.4. 4.2.5.5. 4.3, 4.3.1. 4.3,2. Preventive maintenance conditions. The preventive maintenance condi
23、tions for which the prediction is performed shall be stated in the form of: a) categories and standards of preventive maintenance resources; b) categories of preventive maintenance actions; c) criteria governing the scheduling of preventive maintenance, for example fixed intervals between actions d)
24、 effects on system operational readiness. Corrective maintenance resources. Categories and standards of corrective maintenance resources shall be defined. These may include: a replacement units; b) spare components; c) software media; d) test equipment; e) tools; f) test programs; g) documentation;
25、h) personnel. Maintenance support conditions. The maintenance support conditions for which the prediction is performed shall be in accordance with those stated by the Administration in the equipment/system specification. An a 1 y s i s An analysis has to be made to determine: a) the structure of the
26、 system/equipment; b) the stresses applied to the system/equipment and its parts; c) the maintainability properties of the system/equipment; d) the properties of the maintenance support. Based on this analysis models are built for: - the reliability structure, - the maintainability structure, - the
27、availability structure. The mathematical model used for each characteristic and the derivation of applied formulas shall be stated or referenced. If the prediction is performed by a procedure which preceeds stepwise through several functional levels of the system/equipment, the mathematical models u
28、ses shall be presented separately for each characteristic. Data sources Reliability data. The sources of reliability data shall be agreed by the Administration. Reliability data used, such as failure rates or mean times between failures at unit level, shall be stated. Maintainability data, Maintaina
29、bility data used, such as mean active repair times at different levels, failure detection probability, failure localization efficiency, shall be stated. Maintenance support data. Maintenance support data used, such as numbers of repair men and spare parts, shall be stated either directly or in proba
30、bilistic terms. They shall be consistent with the maintenance support conditions. Prediction results The numerical results shall be clearly presented for each specified characteristic, in correspondence with the corresponding required value. or degree of wear-out; Failure simulation Failure simulati
31、on is merely used for testing maintainability related parameters. Quantity of failures to be simulated The total number of failures to be simulated shall be determined from the adequate test plan (para- graphs 4.4., 4.5. or 4.6.). Distribution of failures For each group of components belonging to a
32、given family (transistor, integrated circuits) and belonging to a given part of the equipment, the number of simulated failures shall be proportional to the mean number of failures which are likely to occur in field operation among the components of this group. When the failure rates of the componen
33、ts are constant, the number of failures to simulate in each component is proportional to its failure rate: the following paragraphs give the details of the method according to this hypothesis. The failure rates to be used in this respect are to be agreed upon by the supplier and by the Administratio
34、n. Edition of June 30, 1989- 1 232bY/Y 000bL7B Y 1. 1 T/N 45-01 E Page 7 2 3 P 4.3.3. Application a) Classify the p parts of the equipment in decreasing order according to the sum A, of the failure rates b) Classify the q families of components in decreasing order according to the sum A, of the fail
35、ure rates c) Give to each group “family of type q belonging to part p” a weight 2 where: of their components. of their components in each family. h A hpp is the sum of the failure rates of the components belonging to the group pq. A is the sum of the failure rates of all the components of the equipm
36、ent. A table as the following one summarises the tasks a), b) and c) above, d) Determine for each group (case pq of the table) the number ii pq of failures to be simulated. If N is the total number of failures to be simulated, n pq is given by: npq = N x h, with A = A, + 4 + . 4 and The computations
37、 generally lead to non-integer values of n pq. These figures will be systematically rounded to the nearest lower integer, thus leading for certain cases to zero. Thus, over the N failures to be simulated, some remain being not assigned: the pq groups in which a remaining failure will be simulated wi
38、ll be chosen at random among those for which n pq C, Le. one concludes that the equipment complies with its specification of failed item localization efficiency, even when the H1 hy- pothesis is true which means that the failed item localization is caracterised by the minimum acceptable values ql an
39、d q2. The risks u and are generally equal. Qimntity of tests to be performed (Under study,) Decision criteria (Under study.) Test of mean durations The statistical test described below does not need any hypothesis on the distributions of the durations. In this test, the mean values Y of n observed d
40、urations is compared to a decision criteria L. One concludes that the true mean value of the considered durations is less than or equal to the value mO proposed by the supplier, as a basis for availability computations, if This test involves two risks of false conclusions : - the risk u of the suppl
41、ier is the probability that Y L, even when the true mean m of the considered durations is equal to the value mO so that one concludes wrongly that the equipment does not comply with the value taken into account in the computations of fiinctional failure rate or of availability; - the risk of the Adm
42、inistration is the probability that Y 5, then n must be increased. If D 3, it is recommended to increase n. If D is larger than the (possibly) specified maximum value, then n should be increased. . 6. DECISION CRITERIA The calculated value C is compared with r, the observed number of failed items or
43、 unsuccessful events. If - r G C, then the specified requirements are regarded as having been complied with - r C, then the specified requirements are regarded as having not been complied with 7. MATHEMATICAL BACKGROUND 7.1 The binomial distribution If the probability of an event is q (approximately
44、 constant), then the probability that the event will occur exactly r times in n observations is: 7.2. where The probability 1 -P(r) of finding r or less events in n observations is: i-a=P(r)=C (3 q i (1 -I i=O For given n, q0 and u nom the C-value is calculated as the lowest integer satisfying The t
45、ables 1, 2 and 3 shown are calculated according to (3) for all n. Approximation formulas A very convenient and rather good approximation for the described test plans is the arcsin transformation (given by R.A. Fisher) for confidence limits for the binomial distribution (ref. 2) with slight modificat
46、ions: For given q0, a and n, the following formula (4) can be applied to find the C value with extremely good approximation for the ranges: 0.001 GqOG0.20 2.5%Ga5 3 I 32 2%; 1 33 221 43 3; 4 :+ 3 5 6 6 6 6 I l 7 7 8 8 8 8 9 9 9 9 9 IO 10 IO IO II 11 II 11 11 12 12 12 12 13 13 13 13 14 14 13 5 5 5 5
47、6 6 6 6 7 7 I l 8 8 8 8 8 9 9 9 9 9 10 IO IO IO IO 11 11 II II 12 12 12 12 12 13 13 - 4 4 4 4 5 5 5 5 6 6 6 6 6 7 I 7 I l 8 8 8 8 8 9 9 9 9 9 IO IO IO IO - fi 3 33 443 444 444 544 544 554 554 555 655 655 665 665 666 766 I66 776 776 176 877 871 877 887 887 988 988 988 998 998 IO 10 9 8 II IO 9 9 II 1
48、0 9 9 11 IO 9 9 11 10 IO 9 11 10 10 9 12 II IO 9 12 11 10 IO n 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 12 13 74 15 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 91 98 99 I O0 C u (Yo) 1.0 2.5 5 IO 14 13 12 11 14 13 12 11 14 13 12 II 15 14 13 12 15 14 13 12 15
49、14 13 12 15 14 13 12 15 14 13 12 16 15 14 12 16 15 14 13 16 15 14 13 16 15 14 13 16 15 14 13 Il 15 14 13 17 16 15 14 17 16 15 14 17 16 15 14 17 16 15 14 18 16 15 14 18 Il 16 14 18 17 16 15 18 17 16 15 18 17 16 15 19 17 16 15 19 18 17 15 19 18 17 15 19 18 17 16 19 18 17 16 20 18 17 16 20 19 17 16 20 19 I8 16 20 19 IS 17 20 19 18 17 21 19 18 17 21 19 18 17 21 20 19 17 21 20 19 17 21 20 19 18 22 20 19 18 22 20 19 I8 22 21 19 18 22 21 20 I8 22 21 20 18 23 21 20 19 23 21 20 19 23 22 20 19 23 22 21 19 23 22 21 19 24 22 21 19 24 22 21 20 15 20 IO IO IO IO II IO II IO