1、DOD-HDBK-771 HI m 9777770 0057778 5 m DOD-HDBK-79 1 (AM) 17 YARCH 1988 MILITARY HANDBOOK MAINTAINABILITY DESIGN TECHNIQUES METRIC NO DELIVERABLE DATA REQUIRED BY THIS DOCUMENT AREA MISC DISTKIBUTION STATEMENT A. APPROVED FOR PUBLIC KELEASE; DISTRIBUTION IS UNLIMITED. Provided by IHSNot for ResaleNo
2、reproduction or networking permitted without license from IHS-,-,-DOD-HDBK-771 NI m 7977770 0057977 7 m DOD-HDBK-791 (AM) DEPARTMENT OF DEFENSE WASHINGTON, DC 20301 Maintainability Design Techniques 1. This standardization handbook was developed by the School of Engineering and to document the requi
3、rement in the Decision Coordinating Paper (DCP) and System Concept Paper (SCP). Maintainability is a risk area not because the require- ments are not technically available; rather, it is a risk area because of the reluctance of the technical community to change from its traditional emphasis on perfo
4、rmance as opposed to maintainability. In summary, maintainability has emerged as an impor- a . a tant factor af the design process and an inherent design characteristic that is truly quantitative in nature and, therefore, lends itself to specification, demonstration, and trade-off analysis with such
5、 characteristics as reliability and logistic support. The implementation of this philos- ophy seeks the goals and objectives presented in par. 1-4.1. For the maintainability engineer this means that the optimum degree of maintainability must be incorporated in system design, beginning as early as th
6、e concept phase. Ifthe maintainability engineer, workingwith the designer, fails to accomplish this, he fails to achieve his objective- i.e., the provision of operational availability. A system that fails to perform at times cannot safely be planned, which renders it useless for combat operations. 1
7、-2 MAINTAINABILITY VS MAINTENANCE Maintainability is a characteristic of design and instal- lation. This characteristic is the measure of the ability of an item to be retained in or restored to a specified condi- tion when maintenance is performed by personnel having specified skill levels and using
8、 prescribed procedures and resources at each prescribed level of repair (Ref. i). Maintenance is essentially the response to the main- tainability program, Le., the series of actions necessary for retaining materiel in or restoring it to a serviceable condition. Maintenance actions are of two types,
9、 i.e., 1. Corrective Maintenance. An action required when equipment fails or malfunctions 2. Preventive Maintenance. An action required to maintain equipment in an operable condition through periodic servicing and/ or replacement of components at specified intervals. Preventive maintenance can, and
10、shouId, be conveniently scheduled to avoid interference with operating schedules. A detailed discussion of preven- tive maintenance is presented in Chapter S. Erroneously, corrective maintenance is referred to as , unscheduled maintenance, and preventive maintenance is referred to as scheduled maint
11、enance. From a practical standpoint military personnel perform maintenance- both corrective and preventive-whenever a window of opportunity exists. The specific maintenance tasks are a function of the reliability, availabilitv. maintainability, 1- I - - Provided by IHSNot for ResaleNo reproduction o
12、r networking permitted without license from IHS-,-,- DOD-HDBK-773 NI m 7777970 0058035 T m DOD-HDBK-791 (AM) and durability (RAM-D) of the equipment and the opera- tional environment, The calendar time, Le., when the maintenance action was performed, makes that action scheduled or unscheduled. The u
13、nscheduled interruption of a planned operation is always undesirable and usually costly; in the extreme case it could be catastrophic. Although unreliability is usually the primary case of failure and thus governs the frequency with which maintenance actions are neoessary, the ease of maintenance an
14、d the skill of maintenance personnel gov- ern the duration of the action. The easier it is to maintain an item of equipment, the fewer will be the demands on both the skill and number of personnel and, in general, the greater the reduction of equipment downtime. Accord- ingly, since the time require
15、d for maintenance actions is a function of the maintainability characteristics of the equipment, effectiveness of built-in testing and physical design features that affect the speed and ease with which maintenance can be performed should be addressed. Design features are discussed in par. 2-5. In ad
16、dition to physical design features, personnel and human factor considerations are of prime importance. These considerations include the experience of the techni- cian, training required, skill level, supervision required, supervision available, techniques used, physical coordi- nation and strength a
17、nd number of technicians, and teamwork requirements. Personnel and human factors are emphasized because the Army-as well as the other Services-is imposing a strength cap on the number of military personnel and restricting the availability of funds for development and training, and procurement of tra
18、in- ing aids. Additionally, the scenario under which the new Army counters threats requires the deployment of light infantry. Thus the impact on engineering design must be that of ease of maintenance, adequate manlmachine interface, minimal maintenance, and maximum surviv- ability. In no single area
19、 of weapon engineering are the potential rewards as great as those which could be achieved by simplifying the human functions needed to maintain the weapon system. This brief introduction highlights the distinction between maintainability and maintenance. In summary, maintainability is a design char
20、acteristic that makes pos- sible the accomplishment of operational objectives with minimal expenditure of support effort and resources and is a prime responsibility of the maintainability engineer working in cooperation with the designer; maintenance is the actions necessary for retaining materiel i
21、n or restoring it to a serviceable condition. 1-3 MEASURES OF MAINTAINABILITY 1-3.1 GENERAL In par. 1-1 it was pointed out that the maintainability characteristic had to be expressed quantitatively to be meaningful. This characteristic is expressed as the proba- 1-2 bility that an item will be retai
22、ned in, or restored to, a specified condition within a given time period if pre- scribed procedures and resources are followed. There are several measurable parameters that can be used to quan- tify the maintainability characteristic, ease of mainte- nance. Ease of maintenance characterizes the main
23、tain- ability designed into an equipment and can be measured by the elapsed time in which the maintenance can be performed. Thus the maintenance time required to cor- rect equipment performance deviations, such as failure or degradation, is a good measure of how well the equipment has been designed
24、for maintainability. When maintenance time as a design parameter is mea- sured, active time only should be considered, The empha- sis is on the word “active” since there are administrative and logistic delays-e,g., absence of proper instructions and waitingfor a repair part-that bear no relationship
25、 to equipment design. Active-type maintenance time for corrective mainte- nance actions usually consists of three sequential steps, i.e., 1. Time to locate the parts requiring repair 2. Time to perform the repair 3. Time to verify that the repair has been performed successfully. For preventive-type
26、maintenance, the first step is elimi- nated because the equipment maintenance area is predetermined. Attributes of the equipment that cause variations in repair time result from the physical characteristics of the failed parts, their location and mounting arrangements in the equipment, and thus thei
27、r accessibility and replace- ability. Variations in the diagnostic time result from trou- bleshooting procedures, location of test points and kind of test equipment, and the sequence in which trouble- shooting is performed. Also technicians skills, their degree of familiarity with the equipment, and
28、 the environ- ment in which maintenance is being performed affect both diagnostic time and repair time. Even identical maintenance actions, caused by identical failures in iden- tical equipment and performed by the same repairman or repair crew, will have varying maintenance times. A large number of
29、 measurements probably would yield a contin- uous distribution of maintenance time for this mainte- nance action even if it were performed by the same techni- cian or the same repair crew. The distribution of mainte- nance times for the same maintenance action with a different repair crew would also
30、 be different. Fig. 1-1 illustrates this phenomenon for two different crews. Crew 1 performs the maintenance action with a mean time rnl ; Crew 2 takes a considerably longer mean time m2. Also the variance of maintenance time with which Crew 2 works is substantially greater than that of Crew 1. Assu
31、me that both crews performed the identical mainte- nance operation under identical environmental condi- Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-DOD-HDBK-771 NI 7777770 OOSOLb L m I I I I I I I I I I I I DOD-HOB K-791 (AM) m rCrew2 I I I 1 111
32、, % Maintenance Timet, h Figure 1-1. Distribution of Maintenance Time tions, at the same time of day, and are refreshed. It is reasonable to conclude from Fig. 1-1 that Crew 1 is more skilled and perhaps more disciplined than Crew 2. Obviously, the time required for a given maintenance action is not
33、 a fixed value. This variability would be even more evident in sophisticated materiel where many differ- ent types of failures may occur, and where the necessary mainienance actions have their own time distributions and usually occur at a different rate. The previous discussion introduced terms-dist
34、ribu- tion, mean, and variance-used by the statistician. This suggests that maintenance time can be defined and ana- lyzed by rigorous statistical techniques. To employ these techniques, specific elements describing maintenance time are necessary-a probability density function g(t) that shows the fr
35、equencies with which maintenance times of different duration occur and a cumulative probability distribution M() that shows the probability that mainte- nance time is equal to or shorter than a fixed time con- straint. Fig. 1-2 represents the case for which the repair time appears to be normally dis
36、tributed. The probability density function presented in Fig. 1-2(A) shows the fre- quency of maintenance occurrence versus a maintenance time; the cumulative probability distribution, Fig. I-2(B), shows the probability of accomplishing maintenancever- sus maintenance time. Some points of interest on
37、 the time axis of Fin. 1-2 a should be noted. Mrepresents the arithmetic meanof the a maintenance time distribution and, in this special case, it also represents the median of the distribution since a Gaussian distribution is indicated. Mmox is the 95th per- centile and represents the time in which
38、at least 95% of all maintenance actions can be compIeted. M(t) is obtained by the integrafion of g(t). This is seen in Fig. 1-2 where the probability M(t) on the maintain- ability axis of Fig. I-2(B) corresponds to the area M(T) under the probability density curve of Fig. 1-2(A). M( 7) is the probab
39、ility that maintenance can be performed in a time Tor less. Ref. 2 provides a detailed discussion of these relationships and the statistical concepts. Since M(t) is a probability, it can assume only positive values, Le., O I probability I f. A probability of zero means impossi- bility; a probability
40、 of one means certainty. Because of its simplicity, the Gaussian case was used to illustrate the meaning of the statistical terms distribution, mean, and variance. In practice, however, a lognormal distribution frequently is assumed to describe the distri- bution of maintenance times. If tractabilit
41、y is desired, an exponential distribution may be assumed. Techniques such as the Kolmogorov-Smirnov test, which tests the hypothesis that the maintenance time was drawn from a population having a probabiiity density function g(t), should be used to determine the best-fit distribution before proceedi
42、ng to compute mean times to repair. Ref, 3 describes the lognormal distribution and the Kolmo- gorov-Smirnov test. 1-3 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-DOD-HDBK-771 HI m 7777770 0058037 3 m DO D-H DB K-791 (AM) Active Maintenance Time
43、t. h (A) Probability Densitv Func!ion I I I I I I O I 9 u c O.sr- I 3 2 M(T) - 7 2.1 Mm, Actin, MaintI x minutes. 1. Part configuration 2. Maintenance scheduling - 3. Simplified diagnostic techniques. 2-2.6.1 Part Configuration Maintainability simplification can be compared to a successful “do-it-yo
44、urself” kit. A good design is so simpli- 2-3 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-DOD-HDBK-791 NI 7979770 0058025 2 m DOB-HDBK-791 (AM) fied that a diagram of the item immediately suggests the method of assembly or disassembly; however, un
45、til a technician is thoroughly familiar with the operation, he should be encouraged to study the documented procedure- the “read instructions when everything else fails” is a dangerous philosophy. In a design of this nature, assem- bly is obvious by means of component parts that fit together in a un
46、ique manner because of their external configuration. Also since familiarity and use of a com- mon part are important considerations in part design and selection, maximum effort should be made to promote the use of standard parts and components. The assembly of parts can be simplified by part design
47、or by support equipment designed to position parts so that they cannot be incorrectly aligned or mounted. A common method is to provide locating holes for the inser- tion of dowel pins to position parts in thecorrect orienta- tion for mounting or for ease of insertion. Support eqGpment, by proper de
48、sign, can permit ease of position- ing for installation of attachments-e.g., a carrier for a weapon pod could be designed with a wing-to-carrier ,guidebar that could align the pod for quick attachment. Removable fasteners, e.g., screws and bolts, should be of the same size wherever possible and of a
49、 standard size, and only in the number required to secure an item prop- erly. This simple design feature will reduce the parts inventory and lessen the requirement for special tools. 2-2.6.2 Scheduling of Maintenance Scheduled and deferrable maintenance functions can often be performed during the same time period, which results in reduced equipment downtime, i.e., increased availability. For example, a modification work order , (MWO) that does not affect system safety o
