Nowlan and Heape RCM Study (US DoD funded)

2,819 views

Published on

This book explains basic concepts, principles, definitions, and applications of a logical discipline for development of efficient scheduled (preventive) maintenance programs for complex equipment, and the on-going management of such programs. Such programs are called reliability-centered maintenance (RCM) programs because they are centered on achieving the inherent safety and reliability capabilities of equipment at a minimum cost. A U.S. Department of Defense objective in sponsoring preparation of this document was that it serve as a guide for application to a wide range of different types of military equipment.

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
2,819
On SlideShare
0
From Embeds
0
Number of Embeds
590
Actions
Shares
0
Downloads
83
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Nowlan and Heape RCM Study (US DoD funded)

  1. 1. Unclassified - - , -, A.,<,C I %TON r , < m ,. r , < ! l h ~ ! #> a a r z > I j r*.r<.j , >% " h REPORT DOCUMENTATION _..- _ PAGE .__ K E A I ) INS I R U C T I O N S I ~ K I , O I + I . :r o h ! p l . : ; m s r . _ pore.! _ I _. : GOV- ACZEl;lOk . . . . hO 3 . H t C I P l C l . T S C A T A L O G NIIIARFH . . ~ --I I ... - - ~ .- ~ i 5 TYPE OF RFPORT d FLHlOD COILIIF:: I, . ~Final ~ - Reliability-Centered Maintenance 5 ~ I J k l l F O H M I N . j OLi, NL"?I11 ~ tiiiM*kr( I F. Stanley Nowlan I MDA 903-75-C-0349 Howard F. Heap-- - .~ . . :)l<rAN(* ~ P I,: irs I U U h U C N N b M I AND A D D R E 5 5 United Airlines San Francisco International Airport San Francisco, Ca 94128 ..-- . .. . ~- .r .i p . 4 ~ 3 ~ I.: a : : ~ : . . . . - -. .-... .-.. 1 ./ w - ~ ~ . . -. - :~) &. - c~ - - - - Office of Assistant Secretary of Defense / December 29,1978 (Manpower, Reserve Affairs and I ogistics) 1 *, ,, ,..,,,F~, ~ ~ d~ ,a , t - Washington, DC 20301 I 476 .- .. . .. .4 ~. -~ 4 .1. .. - i i ( t i . :. .L*. v *YI 3 4 ~ 7 . 1 ~ 35 , 1 !,I: r.m* rrv:~. t . m . r . Iltn. UIIIII- IS / ,L.(.""IT, - - -~ ,.LC,% ,.I ,I., rrli r I :!nclassified q p r o v e d for public release; distribution unlimited. Copies may be obtained from thc National Tcchnical Information Service or !he Defense Documentation Center. - I - .. .. - 6.1 i i7 This work was performed by United Airlines under the sponsorship of the Office of !.ssistant Secretary of Defense (Manpower, Reserve Affairs and Logistics) ., _ ( .* , ,,,.,.. 1 . r.> ..,.. ,,. ., .,, . . < ~ ,,., ,,., , , ,,, ,,, ,,,. !,,.,,, ..,,,, *<.. Actu, ral analysis Aircraft maintenance Cost effectiveness Decision theory Evaluation Failurt effects Flight safety Logistics Maintenance program Mechanical safety Preventive maintenance Product in~provement Reliability Scheduled niaintenance System effectiveness - - . . .. . .. -. . ., ., .,, r , , , , , i . . . r - ~- . - ,,.,.,. .> r . . . .,. ~ ., 11 , I ... r. ,,,. r , , ~ I . . .,I, ~ ~ This book explains basic concepts, llrinciples, definitions, and applications of a logical discipline for devel~pment efficient scheduled (preventive) maintenance of programs for complex equipment, and the on-going management of such programs. Such programs are called reliability-centered maintenance (RCM) programs because they are centered on achieving the inherent safety and reliability capabilities ofL - - - I,-)D ,,,:", 1473 , 9 ; , ! . 8 % , .:.I -, Unclassified -. <&, . . ~~ . ~ , ~ , , . ~ c , A, ; - 1 , ) ~ ,I Tbdir, V l i . 1 . l l l l l l r l l,ntn I-r>ll.rer! EPRODVCtD BY NATIONAL TECHNICAL INFORMATION SERVICE I U.S. DLPARlMtNl OF COMMERCL SPR:NGFltLD, VA. n 6 11
  2. 2. equipment at a minimum cost. A U.S.Department of Defense objective in sponsor- ing preparation of this document was that it serve as a guide for application to a wide range of different types of military equipment. There are essentially only four types of tasks in a scheduled maintenance program. Mechanics can be asked to: Inspect an item to detect a potential failure Rework an item before a ~neximumpermissible age is exceeded Discard an item before a maximum permissible age is exceeded Inspect an item to find failures that have already occurred but wert not evident to the equipment operating crew A central problem addressed in this book is how to determine which types of sched- uled maintenance tasks, if any, should be applied to an item and how frequently assigned tasks should be accomplished. The use of a decision diagram as an aid in this analysis is illustrated. The net result is a structured, ~ystematicblend of experience, judgll~ent,and operational datalinformation to identify and analyze which type of maintenance task is both applicable and effective for each significant item as it relates to a particular iype of equipment. A concludingchapteremphasizes the key importance of having a mutually supportive partnership between the per- sonnel responsible for equipment design and the personnel responsible for equip- ment maintenance if maximum HCM results are to be achieved.I Appendices are included as follows:I Procedures fur auditing ihe development and implementation of a;.c RCM program A historic~l review of equipment maintenance evolution Techniques of performing actuarial analyses An annotated bibliography
  3. 3. RELIABILITY-CENTERED MAlNTENANCE
  4. 4. reliability-centered
  5. 5. maintenance F. STANLEY NOWLAN HOWARD F. HEAP
  6. 6. Prodc~cedby Dolby Access PressDesigner Nancy Clark, illustrators David A. Strassmanand Evanell Towne, compositor Jonathan D. Peck,printer Braun-Brum.field, Inc.
  7. 7. prefacemls VOLUME provides the first full discl~ssionof reliability-centeredmaintenance as a logic~l discipline for the development of scheduled-mairctenafice programs. The objective of such programs is to realize theinherent reliability capabilities of the equipment for which they aredesigned, and to do so at minimum cost. Each scheduled-maintenancetask in an RCM program is generated for an identifiable and explicitreason. The consequences of each failure possibility are eva!uated, andihe failures are then classified according to the sevcrity of their conse-quences. Then for all significant items- those whose failure involvesoperating safety or has major economic consequences- proposed tasksare evaluated according to specific criteria of applicability and effective-ness. The resulting scheduled-maintenance program thus includes allthe tasks neccssay to protect safety and operating reliability, and onlythe tasks that will accomplish this objective. Up to this point the only document describing tllr nse of decisiondiagrams for developing maintenance programs has becn MLiG-2, thepredecessor of P.CM anillysis. MSG-2 was concerned primarily with t l ~ edevelopment of prior-to-service programs and did not cover the us€ 9 foperating information to modify the maintenance program after theequipment enters service or the role of product irnprovement in i;uip-ment development. The chief focus was on the identificatiorl of a set oft a s ~ that would eliminate the cost of unnecessary maintenance without scompromising safety or operating capability. There was no mention ofthe problem of estabiishing task intervals, of consolidating the tasksinto work packages, or of making decisions /here the necessary infoi-mation is unavailab!~.Thc treatment of structure programs was sketchy,and zonal and other general inspection programs were not discussedat all. vii
  8. 8. The difficulty that many people experienced in attempting to apply the concepts of MSG-2 indicated the need for change.; and additions simply to clarify many of the points. It was also abundantly clear, hi .w- ever, that the scope of the material should be expanded to cover the topics that had not been hiccussed in that document, This volsme includes a major expansion of the discussion on the problem of identifying func- tionally and stn~cturallysignificant items. The RCM decision diagram itself is quite different from the one used for MSG-2. Instead of beginning with the evaluation of proposed maintenance tasks, the decision logic begins with the factor that determines the maintenance requirements of each item - the consequences of a functional failure- and then an evalu- ation of the failure modes that cause it. Tllis new diagram also recog- nizes the four basic maintenance tasks that mechanics can perform (instead of three maintenance processes), thereby clarifying the treat- ment of items with hidden functions. The role of a hidden-function failure in a sequence of nw!tiple independent failures is stressed, and it is also shown that the consequences of a possible multiple failure are explicitly recognized in the definition of tl-,e consequences of the first failure. Another important aspect of the RCM decision logic is that it includes a deiault strategy for making initial maintenance decisions in the absence of full information. There is a full discussion of the problem of assigning task intervals, particular!^ those for first and repeat on- condition inspections. The role of age exploration and the use of infor- mation derived from operating experience, both to modify the initial maintenance program and to initiate product improvement, is discussed at ler,gth. The content of scheduled-maintenance programs developed by experienced practitioners of MSG-2 techniques may be quite similar to the programs resulting from RCM analysis, but the RCM approach is Inure rigorous, and there should be much more contidence in its out- come. The RCM technique can also be learned more quickly hnd is more readily applicable to complex equipment other than transport aircraft. Part Qne of this volume presents a full explanation of the theory and principles of reiability-centered maintenance, including a discus- sion of me failure process, the criteria for each of the four basic tasks, the use of the decision logic to develop an initial program, and the age-exploration activities that result in a continuing evolution of this program after the equipment enters service. Part Two describes the app!ication of these principles to the analysis of typical items in the systems, powerplant, and stn~cture division of an airplane; the consid- erations in packaging the RCM tasks, along with other scheduled tasks, for actual implementation; a the information systems necessary for management of the ongoing maintenance program. The concluding chapter discusses the relationship of scheduled maintenance to operat-viii ing safcty, the desigr-maintenance partnership, and tks application of
  9. 9. KCM analysis both tn in-service fleets and to other types of complexequipment. The text is followed by four appendices. Appendix A outlines theprinciples of auditing a program-development project and discussessome of the common probiems that arise during analysis. This materialprovides an excellent check list for the analyst as well as the auditor andshould be especially useful as a teaching aid for those conductii~g train-ing groups in RCM methods. Appendix B is a historical review of thechanges in maintenance thinking in the airline industry. Appendix C isa discussion of the engineering procedures and techniques used inactuarial analysis of reliability data. Appendix D, written by Dr. JamesL. Dolby, is a discussion of the literature in reliability theory, informationscience, decision analysis, and other areas related to RCM analysis andprovides an annotated guide to this literature as well as to the specificliterature on reliability-centered maintenance. Dr. Howard L. Resnikoffhas written an accompanying mathematical treatment of the subject,titled Matllernaticnl Aspects of Reliability-Centered Maintenance. A book of this nature is the result of many efforts, only a few ofwhich can be acknowledged here. First of all, we wish to express ourgratitude to the late W. C. Mentzer, who directed ,he pioneering studiesof maintenance policy at United Airlines, and to the Federal AviationAdministration for creating the environment in which this work wasdeveloped over the last twenty years. We also thank Charles S, Smithand Joseph C. Saia of the Department of Defense, who defined the ron-tent of the present text and counseled us throughout its preparation.James L. Dolby of San Jose State University, in addition to preparingthe bibliography, contributed his expertise to the text. In particular,he helped to develop the concept of partitioning to identify significantitems and the concept of default answers as part of the decision logic,as well as advising us on the actuarial appendix. Nancy Clark editedour eiforts and organized them for clear exposition. Her logical thoughtprocesses resulted in numerous major improvements throughout andmade possible the successful translation of our r~anuscript textbooktoform. Much help on specific areas of the text has come from friends andcowo~kersin the industry. We especially wish to thank Me1 Stone ofDouglas Aircraft for his extensive help with the structure chapter, JohnF. McDonald of the Flying Tiger Line for his comments on the theoreticalchapters, and John F. Pirtle of General Electric for his comments on thepowerplant chapter. Of the many others whose contributions influencedthe text in some i m ~ o r t a n respect, we give particular thanks to Thomas tM. Edwards of United Airlines, Thomas D. Matteson of United Airlines,Ernest Boyer of the Federal Aviation Adr,linistration, Captain L. Ebbertof the U.S. Navy, Edward L. Thomas of the Air Transport Association,and Robert Card of the University of Missouri.
  10. 10. We are also grateful to the ma; y people at United Airlines who pro-v i d ~ d with specific help and assistance. The manuscript itself would usnot have materialized without the effortsof MarieTilson, who cheerfullytyped and retyped the material through many drafts, We also thankClaudia Tracy, whose artwork made the draft manuscript more readable,and J. Douglas Burch, whose efforts throughout the project helped bringit to completio~~. Finally, we would like to thank the management ofUnited Airlines for its patience m d our wives for their encouragementover the many long months of authorship and publication. F. Stnnley N o w l a r ~ Hownrd F. Heap
  11. 11. contents PREFACE rrii 4 MAINTENANCE PHILOSOPHY .vrliCHAPTER ONE rcm: a maintenance discipline 2 .I I THE EVOLUTION OF RCM ANALYSIS 11 - 2 THE BASIS O F RCM DECISION LOGIC hI.3 RELIABILITY PRCBLFMS I N CCMPLEX EQUIPMENT 01 - 4 A N OVERVIEW .OF MAINTENANCE ACTIVITY II PART ONE THEORY AND PRINCIPLES 15CHAPTER TWO the nature ot fdlure 162.1 M E DEFINITION OF FAFAILURE 172.2 THE DETECTION O f FAILURES ?O2.3 THE CONSEQUENCES OF FAILURE 252.4 MULTIPLE FAILURES 282.5 THE FArlURE PROCLSS 312.6 FAILURE I N COMPLEX ITEMS 37
  12. 12. Z.? QU~NTIT,TWE DfiCRlPTION5 OF FAILURE 392.8 AGE-RUIABILIN CHARACTERISTICS 15CHAP~TRTHREE the four basic rimalntenance tasks 503 1 SCHEDULED 0N.CONDlTION TASKS 513-2 SCHEDULED REWORK TASKS 5h3.3 SCHEDULED DISCARD TASKS 583.4 9CHEDULED FAILURE-FINDING TASKS hl3.5 CHARACTERlSnCS OF THE CASlC TASKS 653.6 THE DIMENSIONS OF A SCHEDULED-MAINTENANCE PROGRAM 713.1 PRODUCT IMPROVEMENT AS PREVENTIVE MAINTENANCE 75CHAPTER FOUR developing the initid program 784.1 THE N A N R E OF SIGNIFICANT ITEMS HO4.2 THE RCM DEC:510N PROCESS 864 a3 USE O THE RCM 9ECISION DIAGRAM F 914.4 DETEDMlYlNG COST E5FECWENESS 1004.5 AGE WPLORATION l!lh4.6 PACKAGING THE MAINTENANCE TASKS 109CtiArnER FIVE evolution of the rcm prqram 1125.1 T)IL USES OF OPERATING DATA 1135.2 REACTING T O StRIOUS FAILURES IIf,5.3 REFINING TWL MAINTENANCE P R O C M M 1215.4 REVISIOYS I N MAINTENANCE REQUIREMENTS 12h5.5 THE PRODUCT-IMPROVEMENT PROCESS 1285.6 RCM PROGRIMS FOR !N-SERVICE EQUIPMENT 137
  13. 13. PART TWO APPLlCATlONS 139CHA~ERslr apptyin~ rcm theory to aircraft 1406.1 A SUMMARY OF RCM PRINCIPLES 1116.2 O R G A N l U l l O N OF THE PROGRAM-DEVELOPMENT T W 1456 - 3 BEGlNhlNG THE DECISION PROCESS 1476 - 4 THE INFORMATION FLOW I N DECISION MAKING 153CHAPTER SEVEN acm analysis of systems 1587- 1 CHARACTERISTICS OF SYSTEMS ITEMS 1597.2 ASSEMBLING THE REQUIRED INFORMATION 1617.3 ANALYSIS OF TYPICAL SYSTEMS ITEMS lhb7.4 ESTABLISHING TASK INTERVALS 192CHAPTER EIGHT rcm analysis of powerplants 1948.1 CHARACTERISTICS OF POWERPIANT ITEM5 1958.2 ASSEMBLING THE REQUIRED INFORMATION 19)8.3 FAILURES OF THE BASIC ENGINE FUNCTION 2058.4 FAILURES OF SECONDARY ENGINE TUNCTIONS 2178.5 M E ROLE OF AGE EXPLORATION 224CHAPTER NINE rcm analysis o structures f 2289.1 CHARACTERISTICS OF STRUCTURAL ITEMS 2299.2 THE STRUCTURAL lHSPECTlON P U N 2389 - 3 ASSEMBLING THE kEQUlRED INFORMATION 2479.4 RCM ANALYSIS OF STRUCTURAL ITEMS 2529 5 ESTABLISHING INITIAL INSPECTION INTERVALS 2509 - 6 STRUCTURAL AGE EXPLORATION 273
  14. 14. CHAPTER TEN completingthe maintenance program 27610.1 OTHER SCHEDULED-MAINTENANCE TASKS 27710.2 PACKAGING THE MAINTENANCE WORKLOAD 284CHAMER ELEVEN the use of operating information 2921 1.1 TYPICAL INFORMATION SYSTEMS 2931 1 - 2 TYPICAL TYPES OF ROUTINE ANALYSIS Xll1 1 - 3 MODIFYING THE MAINTENANCE PROGRAM 3071 1 - 4 INTERVALS: A N INFORMATION PROGRAM 32111 - 5 RESOLVING DIFFERENCES OF OPINION 3251 1.6 PURGING THE PROGRAM 328CHAPTER TWELVE the role of scheduled maintenance 33012.1 SAFETY. RELIABILITY, AND SCHEDULED MAINTENANCE 33112.2 AIR-TRANSPORT SAFETY LEVELS 33712.3 THE DESIGN-MAINTENANCE PARTNERSHIP 34112.4 RCM PROGRAMS FOR IN-SERVICE FLEETS 3431 2 - 5 UPANSION OF RCM APPLICATIONS 317 PART THREE APPENDICES 349APPEND~XA auditin3 rcm pro3ramdevelopment 350A. 1 AUDITING THE PROGRAM-DEVELOPMENTPROJECT 351As2 AUDITING THE DECISION PRCCESS 354A.3 AUDITING ANALYSIS O r THE CQUIPMENT 3h?As4 AUDITING THE ONGOING PROGRAM 3h7A.5 AUDITING NEW PROGRAMS FOR IN-SERVICL FLEETS 3bH
  15. 15. APPENP~X B the history sf rcm programs 370B. I THE HARD-TIME PARADOX 3718.2 CHANGING PERCLPTIONS OF THE HARD-TlME POLICY 376B.3 THE INTRODUCTION OF GN-CONDITC3N MAINTEtiANCE 383B.4 THE AIR TRANSPORT ASSOCIATION MSG- 1 AND MSG-2 PROGRAMS 385B.5 THE RELATIONSHIP OF SCHEDULED MAINTENANCE TO OPERATING SAFETY 387APPENDIX C i I c t ~ a hilfd s t l Wb ! 390C *1 ANALYSIS OF LIFE-TEST DATA 391C.2 ANALYSIS OF DATA FROM A DEFINED CALENDAR PERIOD 395C.3 THE SMOOTHING PROBLLU 402C.4 ANALYSIS OF A MIXED POPULATION 40RC.5 USEFUL PROBABILITY DlSTRIBUTIONS 411C.6 A SPECIAL USE OF THE EXPONENTIAL D:STRIBUTION 417APPENDIX D bibliqraptuy 420D. 1 HISTORICAL DEVELOPMENT 422D.2 RELIABILITY I LORY AND ANALYSIS 425D.3 INFOllMATlON SCIENCL AND DECISION ANALYSIS 427D.4 MAINTENANCE THEORY AND PHILOSOPHY 430D.5 MAINTENANCE APPLICATIOHS 436D.6 A GUlDE TO OTHER SOURCES 47<; GLOSZ4RY 453 INDEX 467
  16. 16. a maintenance philosophy A,. operators maintenance program has four objectives: b To ensure realization of the inherent safety and reliability levels of the equipment To restore safety and reliability to their inherent levels when deteri- oration has occurred * To obtain the information necessary for design improvement of those items whose inherent reliability proves inadequate b To accomplish these goals at a minimum total cost, including main- tenance costs and the costs of residual failures Relialirlity-centered maintenance is based on the following precepts: A failure is an unsatisfactory condition. There are two types of fail- ures: frorctionnl frrilurcs, usually reported by operating crews, and pote!~tinlfnilrrrc.~, usually discovered by maintenance crews. b The consequences of a functicnal failure determine the priority of maintenance effort. These consequences fall into four categories: b saiety consequelrces, involving possible loss of the equipment and its occupants b Operational consequences, which involve an indirect economic loss as well as the direct cost of repair b Nonoperational consequences, which involve only the direct cost of repair b Hidden-failure consequences, which involve exposure to a pos- sible multiple failure as a result of the undetecteci failure of a hidden function
  17. 17. b Scheduled maintenance is required for any item whose loss of func- tion or mode of failure could have safety consequences. If preven- tive tasks cannot reduce the risk of such failures to a n acceptable level, the item must be redesigned to alter its failure consequences.b Scheduled maintenance is required ior any item whose iunctional failure will not be evident to the operating crew, and therefore reported far corrective action. In all other cases the consequences of failure are economic, and maintenance tasks direzted at preventing such failu~esmust be justified on economic grounds.I. All failure consequences, including economic consequences, are established by the design characteristics of the equipment and can be altered only by basic changes in the design: b Safety consequences can in nearly all cases be reduced to eco- nomic consequences by the use of redundancy. b Hidden functions can usually be made evident by instrumen- tation or other design features. b The feasibility and cost effectiveness of scheduled main- tenance depend on the inspectability of the item, and the cost of corrective maintenance depends on its failure modes and inherent reliability.* The inherent reliability of the equiprnent is the level of reliability achieved with an effective maintenance pr0gram.Thi.s lc.!el is estab- lished by the design of each item and the manufacturing processes that produced it. Scheduled maintenance can ensure that the in- herent rcliability of each item is achieved, but no form of mainte- xvii
  18. 18. nance can yield a level of reliability beyond that inherent in t4e design. A reliability-centered mainteqance program includes only those tasks which satisfy the criteria for both appIicability and effectiveness. The opplicilbility of a task is determined by the characteristics of the itcin, and its effective?rcssis defined in teims of the consequences the task is designed to prevent. There are four basic types of tasks that mechanics can perform, each of which is applicable under a ilnique set of conditions. The first three tasks are directed at preventing functional failures of the items to which they are assigiled and the fourth is directed at pre- venting a multiple failure involving that item: b On-condition inspections of an item to find and correct any potential failures b Rework (overhaul) of an item at or before some specified age limit b Discard of an item (or one of its parts) at or before some speci- fied life limit b Failure-finding inspections of a hidden-function item to find and correct functional failures that have already occurred but were not evident to the operating crew A simple itenr, one that is subject to only one or a very few failure modcs, frequently shows a decrease in reliability with increasing operating age. An age limit may be useful in reducing the overall failure rate of such items, and safe-life limits imposed on a single part play a crucial role in controlling critical failures. A co!;rple.~it~irr, one whose functional failure may result from many different failure modes, shows little or no decrease in overall reliability with increasing age unless there is a dominant failure mode. Age limits imposed on complex t.omponents and systems (including the equipment itself) therefore have little or no effect on their overall failure rates. The RCM decision diagram provides a logical tool for determining which scheduled tasks are either necessary or desirable to protect the safety and operating capability of the equipment. The resulting set of RCM tasks is based on the following considera- tions:xviii b The consequences of each type of functional failure
  19. 19. b The visibility of a function..l failure to the operating crew (evidence that a failure has occurred) b The visibility of reduccii r~sistanceto failure (evidence that a lailure is imminent) b The age-reliability characteristics of each item b The economic tradeoff between the cost of scheduled main- tenance and the benefits to be derived from itb A m~ltiple failure, resulting from a sequence of jndependent fail- ures, may have consequences that wo%ld be cacsed by any one not of the individual failures alone. These consequences are taken into account in the definition of the failure consequences for the first failure.b A default strategy governs decision making in the absence of full information or agreement. This strategy provides for conservative initial decisions, to be revised on the basis of information derived from operating experience..4 scheduled-maintenance program must be dynamic. Any prior-toservice program is based on limited information, and the operatingorganization must be prepared to collect and respol~dto real datathroughout the operating life of the equipment.b Management of the ongoing maintenance program requires an organized information system for surveillance and analysis of the performance of each item under actual operating conditions. This information is needed for two purposes: b To determine the refinements and modifications to be made in the initial maintenance program (including the adjustment of task intervals) b To determine the needs for product improvementb The information derived from operdting experience has the follow- ing hierarchy of importance: b Failures that could affect operating safety b Faiiures that have operational consequences b The failure modes of units removed as a result of failures b The general condition of unfailed parts in units that have failed b The general condition of serviceable units inspected as samples xix I
  20. 20. b At the time an initial program is developed information is dvailable to determine the tasks necessaiy to protect safet;, and operating capability. However, the infomatior, reqxired to determine oyfi- mum task intervals and the applicability of age limits can be obt~inedonly from age exploration after the equipment enters service.b With any new equipment there is always the possibility of un- anticipated failure modes. The first occurrence of any serious unanticipated failure immediately sets in motion the followicg product-improvcrnent cycle: b An on-condition task is developed to prevent recurrences while the item is being redesigned. b The operating fleet is modified to incorporate the redesigned part. b After the modification has proved successful, the special task is eliminated from the maintenance program.b Product improvement, based on identification of the actual relia- bility characteristics of each item through age exploration, is part of the normal development cycle of all complex equipment.
  21. 21. RELIABILITY-CENTERED MAINTENANCE
  22. 22. CHAPTER OtiEreliability-centered maintenance m TERM r e l i l a l i l - c t r c rrrnirrtc~rm~r(-c E refers to a scheduled-maintenance program designed to realize the inherent reliability capabilities of equip- ment. For years maintenaxe was a craft learned thrwgh experience and rarely examined analytically. As new performance requirements led to increasingly complex equipment, however, maintenance costs grew accordingly. By the late 1950s the vi::ume of these costs in thtr air- line industry had reached a level that warranted a new look at the entire concept of preventive maintenance. By that time studies of actual oper- ating data had also begun to contradict certain basic assumptior~sof traditional maintenance practice. One of the underlying assumptions of maintenance theory has always been that there is a fundamental cause-and-effect relationship between scheduled mairpnance and operating reliability. This assump- tion was based on th .,tuitive belief that because mechanical parts wear out, the reliability 01any equipment is directly related to operating age. It therefore followed that the more frequently equipment was over- hauled, the better protected it was against the likelihood of failure. The only problem was in determining w!,at age limit was necessary to assure reliable operation. In the case of aircraft it was also commonly assumed that all reli- ability proljlems were directly related to operating safety. Over the years, however, it was found that many types of failures couid not be pr2vented no matter how intensive the maintenance activities. More- over, in a field subioct to rapidly expanding technology it was becoming increasingly difficult to eliminate uncertainty. Equipment designers2 INTRODUCTION were ible to cope with this problem, not by preventing failures, hut by
  23. 23. preventing such failures from affecting safety. In most aircraft all essen-tial functions are protected by redundancy features which ensure that,in the event of a fdilure, the necessary f u ~ ~ c t i o n still be available willfrom some other source. Although fail-safe and "failure-tolerant" de-sign practices have not entirely eliminated the relationship betweensafety and reliability, they have dissociated the two issues sufficientlythat their implications for maintenance have become quite different. A major question still remained, however, concerning the relation-ship between scheduled maintenance and reliability. Despite the time-honored belief that reliability was directly related to the inteivalsbetween scheduled overhauls, searching studies based on actuarialanalysis of failure data suggested that the traditional hard-time policieswere, apart from their expense, ineffective in controlling failure rates.This was not because the intervals were not short enough, and surelynot because the teardown inspections were not sufficiently thorough.Rather, it was because, contrary to expectations, for many items thelikelihood of failure did no+ in fact increase with increasing operatingage. Consequently a maintenance policy based exclusively on somemaximum operating age would, no matter what the age limit, have littleor no effect on the failure rate. At the same time the FAA, which is ~.esponsible regulating air- forline maintenance practices, was frustrated by experiences showing thatit was not possible for airlines to control the failure rate of certain typesof engines by any feasible changes in scheduled-overhaul policy. As aresalt, in 1960 a task force was formed, consisting of representativesfrom both the FAA and the airlines, to investigate the capabilities of CHAPTER l 3
  24. 24. scheduled maintenance. ?he work of this group led to an FA.;/Industy Reliability Program, issued in November 1961 he he introduction to that program stated: Tile det.elopment of this program is towards the control of reli- ability through an analysis of the factors that affect reliability and provide a system of actions to improve low reliability levels when they exist. ... In the past, a great deal of emphasis has been placed on the control of overhaul periods to provide a satisfactory level of reliability. After careful study, the Committee is convinced that reliability and overhaul time control are not necessarily directly associated topics; therefore, these s ~ b j e c t are dealt with separately. s Because the prcpulsion system has been the area of greatest con- cern in the recent past, and due t3 powerplant dala being more .dadily available for study, programs are being developed for the propulsion system first as only one system at a time can be success- fully worked out. This approach was a direct challenge to the traditional concept that the length of the inter-val between successive overhauls of an item .. was an important factor in its failure rate. The task force developed a propulsion-system reliability program, and each airline involved in the task force was then huthorized to develop and implement reliability prugrams in the area of maintenance in which it was most interested. During this process a great deal was learned about the conditions that must obtain for scheduled maintenance to be effective.t It was also found that in many cases there was no effective form of scheduled maintenance.I 1 THE EVOLUTION OF RCM ANALYSlS t ~1 1 i tt At United Airlines an effort was made ;o coordinate what had been .li.~gr.~tnIt.~.l~t~itlut% learned from these various activities ant1 ?eiine a generally appli- . , l > r ~ ~ l l ~ ~ ~ ~ l*l ~h r~, l tll l ~ : 1 ~ , : cable approach to the design of maintena1::e programs. A rudimentary decision-diagram technique was devised in 1965 and was refined over the next few years.+ This technique was eventually embodied in a docu- FAAll~~dltstry Rcliabilit!! Pro~ra111. Federal Aviation Agency, November 7, 1961, p. 1. tHnt~dlloclkfor M a i n t r ~ ~ n t Co~ttrolIly Kciiellility Mctlli~ds,FAA Advisory Circular 120-17, ~cr December 31, 1964. SH. N. Taylor and F. S. Nowlan, Turhine Engine Reliability Program, FAA Maintenance Symposium on Contin~lcdReliability of Transport-type Aimraft Propulsion Systems, Washington, D.C., NovernL1c.r 17-18, 1965. T. D. Matteson and F. S. Nowlan, Current I rends in Airline Maintenance Progmms, AlAA Commercial Aircraft Design and Opera- tions Meeting, Los Angeles, June 12-14, 1967. F. S. Nowlan, Thc Use of Decision Diagrams for Logical Analysis of Maintenancc Programs, United Airlines internal document, August 2, 1967.
  25. 25. nlent published under the title Handbook: Maintetranee Eval~rationandlJro~ram Dtlvelopmenf,generally known as MSG-I.* MSG1 was used byspecial teams of industry and FAA yersor~nel develop the initial pro- tograin issued by the FAA Maintenance Review Board for the Boeing 747.As described by the FAA, these t e a s s t . . . sorted out the potentid maintenance tasks and then evaluated them to determine which must be done for operating safel.y or essential hidden function protection. The remaining potential tasks were evaluated to determine whether they were economically use- ful. These:procedures provide a systematic review of the aircraft design so that, in the absence of real experience, the best [mainte- nance] plocess can be utilized for each component and system.The Boeing 747 maintenance program so developed was the first:attemptto apply reliebility-centered mait;tenance concepts. This program hasbeen successful. Subsequent improvements in the decision-diagram approach ledin 1970 tc a second document, MSG-2; AirlinclMarrrrfncturcr Mni~rtcnnriccProgram Plnnnitlg Documort, which was used to develop the scheduled-maintenance programs for the Lockhred 1011 and the Douglas DC-lo.$These programs have been successful. MSG-2 has also been applied totactical milit*.y aircraft such # i s the McDonnell F4J and the LockheedP-3, and a similar document , 1,epared in Europe was the basis of theinitial scheduled-maintenance programs for such recent aircraft as theAirbus lndustric A-300 and the Concorde. The objective of the techniques outlined by MSG-1 and MSG-:! wastc develop a qcheduled-maintenance program that assured the maxi-mum safety and reliability of whicn the equipment was capable andwould meet this requirement at the lowest cost. As an example of theeconomic benefits achieved with this type of under traditionalmaintenance policies the initial program for the Douglas DC-8 includedscheduled overhaul for 339 items, w h e r ~ a s initial progranl for the theDC-10, based on MSG-2, assigned only seven items to overhaul. One ofthe items 110longer subject to an overhaul limit in the later progr.:! II wasthe turbine engine. Elimination of this scheduled task qot only led tomajor reduct,ons in labor and materials costs, but also reduced the spare-engine inventory required to cover shop activities by more than 50percent. Since engines for larger airplanes now cost upwards of $1million each, this is a respectable saving.747 hlaintcnance Steering Group. 1111r1dl~ooA: Mr~itltctra~l~.r E~~slrrr~tiotr l1ro,.rr~ttrD r l ~ c - l ~ ~ l r - r~tlrlttrc3tlt( M S G - I ) . Air Transport Association, July lo, 1UhA.t f l , r l r m l Azliatiotl Adtttitlistrntiott Crrtificriticlrl l-rt~ccdltres,May 19. 1972, par. 3036.S A i r l i t ~ z l M n t ~ ~ c f r ~ ~ t ~ l t ~ tl ~ t nP~r lo~ ~ n t t P / ~ l l t l ~ ~ l) ~ C I O ~ I M ISI C -: 2 , Air Transport Mt~it t ~r ~ t O ~ ~Association, H & M Subcommittee, Marcy 25, 1970. SECTION 1 . 1 5
  26. 26. As another example, under the initial program developed for the Boeing 747 it took United Airlines only 66,000 manhours on major struc- tural inspections to reach an inspection interval of 20,000 hours. In con- trast, traditional maintenance policies led to an expenditure of over 4 millien manhours before the same interval was attained for sttuctural inspections 011the smaller and less complex Douglas DC-8. Cost reduc- tions on this scale are of obvious importance to any organization responsible for maintaining large fleets of complex equipment. More important, they are achieved with no decrease in the reliability of the equipment; in fact, a clcarer understanding of the failure process has actually improved operating reliability by making it easier to pinpoint signs of an imminent failure. The specific developments that led to RCM concepts as a fundi- mental approach to maintenance p!anning are described in detail in Appendix 8. Although MSG-1and MSG-2were short working papers, intended for use by a small number of people with extensive back- grounds in aircraft maintenance, further clarification of the bae;.c prin- ciples has resulted in a logical discipline that applies to maintenance programs for any complex equipment.1 2 THE BASIS O RCM DECISION LOGIC F (nc naturc o f failure The principles of reliability-centered maintenance stem from a rigorous idztltific.ltion of examination of certain questions that are often taken for granted: significant items ov.ilu.~tic~n failure b How does a failure occur? of COIlLicl)ullll~CL. selection of .lppli1-.lh~t. .inti What are its consequences? rtfcctivc tasks ,hc of age cx,,loration b What good can preventive maintenance do? One of the chief drawbacks of the old hard-time approach to scheduled maintenance is that the resulting teardown inspections provided no real basis for determining when serviceable parts were likely to fail- that is, there was no objective means of identifying reduced resistance to failure. More than any other single factor, recognition of the specific need to identify potential-failure conditions has been responsible for the change from scheduled overhauls to on-condition inspections for siqns of imminent failure. Unfortunately, not all items can be protected by this type of main- tenance task. In some cases the failure mechanism is imperfectly understood, in others it is random, and in yet others the cost of such inspections excecds the benefits they might provide. In fact, preventive maintenance is not possible for many items of modem complex equip- ment. Nor, in all cases, is it necessary. Failures which could jeopardize6 INTRODUCTION the safety of the equipment or its occupants must be prevented. Under
  27. 27. modem design practices, however, very few items fa]! into this category,either because an essential function is provided by more than one sourceor because operating safety i s protected in some other way. Similarly,hidden functions must be protected by schedultd maintenance, bothto ensure their availability and to prevent exposure to the risk of amultiple failure. In all other cases the consequences of failure are economic, and thevalue of preventive maintenance must be measured in economic terms.In some cases these consequences are major, especially if a failureaffects the operational capability of the equipment. Whenever equip-ment must be removed from service to correct a failure, the cost of fail-ure includes that loss of service. Thus if the intended use of the eauip-ment is of significant value, the delay or abandonment of that usewill constitute a significant loss-a fact that must be taken into accountin evaiuating the benefit of preventive maintenance. Other failures willincur only the cost of correction or repair, and such failures may well betolerable, in the sense that it is less expensive to correct them as theyoccur than to invest in the cost of preventing them. In short, the driving element in all maintenance decisions is notthe failure of a given item, but the consequences of that failure for theequipment as a whole. Within this context it is possible to develop anefficient scheduled-maintenance program, subject to the constraints ofsatisfying safety requirements and meeting operational-performancegoals. However, the solution of such an optimization problem requirescertain specific information which is nearly always unavailable at thetime an initial program must be developed. Hence we also need a basicstrategy for decision making which provides for optimum maintenancedecisions, given the information available at the time. The process ofdeveloping an initial RCM program therefore consists of the followingsteps:b Partitioning the equipment into object categories to identify those items that require intensive studyb Identifying significant items, those whose failure would have safety or major economic consequences for the equipment as a whule, and all hidden functions, which require scheduled maintenance regard- less of their significanceb Evaluating the maintenance requirements lor each significant item and hidden function in terms of the failure consequences and select- ing only those tasks which will satisfy these requirementsb Identifying items for u.hich no applicable and effective task can be found and either recommending design changes if safety is involved or assigning no scheduled-maintenance tasks to these items until further information becomes available I-,.
  28. 28. F Selecting conservative initial intervals for each of the included tasks anci grouping the tasks in maintenance packages for applicationb Establishing an age-exploration program to provide the factual information necessary to revise initial decisions The first of these steps is intended, as a pureiy practical matter,to reduce the problem of analysis to manageable size and to focus itaccording to areas of engineering expertise. The next three steps arethe crux of RCM analysis. They involve a specific sequence of decisionquestions, worded to indicate the information required for a yeslnoanswer in each case. Where this information is not available, a defaultanswer specifies the action that will best protect the equipment untilthere is a basis for some other decision. This dezision-diagram tech-nique, described in full in Chapter 4, not only provides an orderly basisfor making decisions with limited information, but also results in a clearaudit trail for later review. In the airline industry all scheduled-maintenance programs are, ofcourse, subject to FAA review and approval. The initial program foreach new type of equipment is promulgated by the FAA MaintecanceReview Board. This document, developed in conference with the equip-ment manufacturers and the purchasing airlines, forms the basis of theinitial program submitted by each airline for FAA approval. Organiza-tions operating other equipment in the civilinn and rnj!itary sphetesmay define their initial maintenance progranls differently, bc! ramecomparable review procedure is usually involved. Because any initlal scheduled-maintenance program must be devel-oped and imylenlented in advance of actual operational data, an im-portant element of RCM programs is age exploratiol~,a procedure forsystematic gathering of the information necessaly to determine theapplicability of some maintenance tasks and evaluate the effectiveness of otl-.ers. As this information accumulates, the same decision diagramprovides a means of revising and refining the initial program. Much ofthis i n f ~ n r ~ a t i o n already available, of course, for equipment that has isbeen in service for some time. Although the specific data needed mayhave to be retrieved from several different information sy%te;is, andthe remaining useful life of the equipment will be a factor in certaindecisions, RCM analysis under these circumsta:..~es result in fewer willdefault decisions, and hence a near-optimum program at the outset.Such programs usually include a larger number of on-condition inspec- tions than the programs arrived at under older policies, and fewer of the scheduled rework tasks which had been included simply because there was no evidence that they should not be done. An effective scheduled-maintenance program will realize all the reliability of which the equipment is capable. However, ;lo form of pre- ventive maintenance can alter characteristics that are inherent in the
  29. 29. . . dcsinn. Thr8residuaI failures that occur after all appl~cable effective and , yrsventivc tasks have been implemented reflect ihe inherrnt capability uf the equipmrnt. and if the resulting level of reliabi!ity is inad-quate. : the onlirc<ourse is engineering redesign. This effort may be ditected a i P sink& component to correct for a dominant failure mode or it may be directed at some characteristic that will make a particular prevcntivc technique feasible. Pmduc! improvement of this kind takes place-rou- tinely during the early years of operation of any complex equipment. Thus,. although reliability-centered maintenance is concerned tn the short m n with tasks based on the actual reliabiliky characteristics of the equipment, it is also concemcd with improvements that will uitimately increase delivered ~eliability. 1 3 ILUMIUTI PROBLWS IN COMPLEX EQUIPMENT Failures are inevitable in any complex equipment, although their con- lrilurr ru*ihilitic* in sequences can be controlled by careful design and effective mainte- "mr" rquipmm pcrtltrmrnrr requirc.mcnt. nance. The reason for this failure incidence is apparent if we consider and ~r,iahlli,v some basic differences between simple and complex equipment. Simple th, lo~e,,I Jr*ijin in rr.duzin): equipment is asked to p r f o n n very few diiterent functions. Such ~ 1 l ~ l ~ ~ ~ " q U ~ ~ equipment therefore consists of only a few systems and assemblies, and thew in turn may be s simple that some are exposed to only one o possible failure mode. In most cases this simplicity extends t the o structur,~! elements as well, and both thc structure and the various items on the equipment are relatively accessible for inspt~tion. As a result, simple equipme!~t certain distinct failure charac- has teristics. Because it is exposed to relatively few failure possibilities, its overall reliability tends to be higher. For the same seasun, these fai!ures tend to be age-related; each type of failure tends to concentrate around some average age, and since only a few Vpes o hilure are involved, f they govern the average age at failure. Howewer. in the absence of redundancy and other protective features, such faiIures mr have fairly iy serious consequences. Thus simple equipment i s often pmtected by "overdesign"; components are heavier and bulkier than nmssary, and familiar materials and processes are used to avoid the uncertainty associated with more comdex high-yerfotmance equipment. All in all. the traditional idea that failures are directly related to , safety and that their likelihood varies directly with age i; often true for simple equipment. In any case, it is fairly easy to make an exhaus- tivestud-y of such equipment to determine its scheduled-maintenance requirements. The situation is quite different with the complex equipment in t t s e today. The general-aviation aircraft of the 1930s usually bad d si ,-pie
  30. 30. nviprcwating engine, a fixed-pitch propeller, fixed land~ng gear, andno wing flaps. The m d e m airyl.mc may have wveral turboprop orturtwict p ~ w ~ r p l a n t wtractable I ~ n d i n g s, gear, mrrvable high-lilt de-vices, an airframe anti-icing system. pressure- atid ttmperature-contmlsystems for the &in, extensive cnnimunicati~ns navigation q u i p - andmcnt, complex cockpit instrumentation, and ccrmplex ancillary systemsto support all thrsc additional items. This increased complexity hasgreatly expanded the safe operational c~pabilityof the aircraft. Theslmplc airplane of the 19.70s was to trips of a few hundredmiles under wasonibfy favorable weather conditions. The higher per-fornitice capability demanded of m d e m equipment, howcvcr, hasgreatly increased not only the numhcr of items that can fail, but thetypes of failurn that can occur. b c h new design of any high-perfcmndnct*cquipmcnt is essentiallyan attempt to make earlier designs technologically obsolete, with Ihurtsual measure of improvcment k i n g potential opratinp, capability(includ~ng operating costs). In other words, this is the operating capa-biliry cxpcctcd in the absence of any failures that might change theci~umstances.The basis for cvilluatin~new aircraft designs usuallyincludes performance factors such as the following. The maximum payload (military or commercial) that can be carried over a given distancc The maximum distance over which a given payload can be carried The minimum 512~ the vchiclt. that can carry a given payload d over a givcn distancc The highest speed that can be attained under defined payl~.adl range conditions Special capabilities, such as the ability to traverse rough terrain, operate frpm short runways, or withstand punishmentIn somc cases these hctors an! weighed against the anticipated directoperating costs (including maintenance costs) associated with attainingsuch capabilities, since a major objective may k to achieve the mini-mum cost per unit of payload transported. In other cases perforrnaocetakes precectence ovcf cost. This is true not on!y of military equipment.but of certain types of civilian equipment, where there is an adtwuatemarket for specialized c.ipability despite its cost. Another aspect of ,performance demands, of course, is the trendtoward incrcwing automation. Examples arc everywhere-automaficflight-contml systems in aircraft, including automatic approach ahdlanding equipment; automatic transmissions in autumobiles; auto-mated traffic-control syslcms for rapid-transit trains; and automaticaperture-setting devices in cameras.
  31. 31. The design oi complex equipment, therefore, is always a tradeoffbetween achieving the required performance capability and acceptablereliability. This tradeoff entails an intentional compromise between thelightness and compactness required for high performance and theweight and bulk required for durability. Thus it is neither econom-ically nor technologically feasible to produce complex equipment thatcan sustain trouble-free operation for an indefinite period of time.Although the reliability of certain items that perform single functionsmay be improving, the number of such items has been vastly multi-plied. It is therefore inevitable that failures will occur-that is, thatcertain parts of the equipment will lose the capability of performingtheir specified functions. Our concern is not with the number of these failures, but with theconsequences of a given failure for the equipment as a whole. Willthe loss of a particular function endanger the equipment or its occu-pants? If not, is it necessary to abort the mission or take the equipmentout of service until repairs call be made? Or can unrestricted operaticncontinue and the repair be deferred to a convenient time and place? Theability to defer failure coilsequences depends largely on the design ofthe equipment. One strategy is the use of redundancy and fail-safeconstruction. Anoiher strategy is failure substitution, the use of a minorfailure to preempt a major one, as in the use of fuses and circuitbreakers. This latter concept extends to maintenance activities in whichpotential failures are used to preempt fi~nctionalfailures. Thus thedesign may include various iqstrumentation to give some warning ofan impending failure or other features which iacilitate inspection forpossible deterioration. All these features actually increase the numberof failure possibilities in the sense that they add more items that couldfail. However, they greatly reduce the consequences of any single failure. 1 4 AN OVERVIEW OF MAINTENANCE ACTIVITYThe activities of a maintenance organizaiion include both the scheduled corrective and scheduledwork that is performed to avoid failures and the corrective work that is maintenance scheduled-maintenanceperformed after failures have occurred. Our present concern is with p,g,mpreventive maintenance, the program of scheduled tasks necessary to maintenance stationsensure safe and reliable operation of the equipment. The complete col- line maintenance and shoplection of these tasks, together with their assigned iilter-- . is termed maintenancethe scheduled vaintenance program. This program includes o. the tasksthat are scheduled in advai~ce-servicing and lubrication, illspection,and scheduled removal and replacement of items on the equipment.Exhibit 1.1 lists some typicai tasks in such a program. In order to accomplish the anticipated corrective and scheduledmaintenance, an operating orgar.ization must establish an overall sup- SECTION 1 * 4 11
  32. 32. EXHIBIT 1.1 I ypical scheduled-maintenance tasks for various itemson aircraft. Some scheduled titsks are performed on the aircraft atli~tc-maintenancestations and ethers are performed at the majornlaititcnanrr base, either as p , ~ rof a larger maintenance package or t-1s part of the shop procedure whenever a failed unit is sent to theniaintenance base for rcpair. nature of item scheduled-maintenance task Fuel-pump assembly On-con.dltion (on aircraft): 60 operating hours (Douglas A41 Inspect filter for contamination on-condition (on aircraft): 1,000 operating hours Inspect drive shaft for spline weat Brake assembly, main landing On-condition (on aircraft): . During ovm@ht tops gear (Douglas DC-10) Inspect brake wear indicatoia and w a l h u n d checks On-condition (in shop): Test Whenever brake asecmbly automatic brake adjuster is in shop rOwUPLANT ITEMS Compmsor rear frame On-condition (on aircraft): 500 flight cycles or phase (General Electric CF6-6) Inspect front flange for cracks check (134 days), whichever t emanating from bolt holes is first Nozzle guide vanes On-condition (on aircraft): l,000 operating h o w (Pratt & Whitney JT8D-7) Perfonn borescope inspection for burning, cracking, or bowing of guide vanes Tenth-stage compressor blades Scheduled rework: 6,000 operating hours (Pratt & Whitney JT4) Shot-peen blade dovetail and apply antigalling compound Stage 3 turbine disk Scheduled diecard: Replace 15,000 flight cycles or 30,000 (Pratt & Whitney JTSD) turbine disk with new part operating h o r n , whichever is first STIUCTUIAL m s Rear spar at bulkhead On-condition (on aircraft): Primly strength-indicator intersection (Douglas DC-10) Inspect specified intersections areas 5,000 operating houn, in zones 531,631,141,142 for internal fuel-tank areas cracks and c o m s i ~ n 20,000 houre Shock strut, main landing On-condition (in shop): Strip 19,500 operating hours gear (Boeing 737) cadmium plate and inspect for cracks and corrosion
  33. 33. port plan which includes the designation of maintenance stations, staff- ing with trained mechanics, provision of specialized testing equipment and parts inventories, and so on. The overall maintenance plan of an airline i:: ?pica1 of that for any transportation system in which each piece of bquipment operates through many stations but has no unique home station. A large proportion of the failures that occur during operation are first observed and reported by the operating crew. Some of these must be corrected after the next landing, and a few are serious enough to require a change in flight plan. The correction of many other failures, however, can be deferred to a convenient time and location. Those line stations with a high exposure to the need for immediate corrective work are designated as maintenance stations and are equipped with trained mechanics, spare-parts inventory, and the facilities necessary to carry out such repairs. United Airlines serves 91 airline stations with 19 such maintenance stations. The decision to designate a particular station as a maintenance station depends chiefly on the amount of traffic at that station and the reliabiliv of the aircraft involved. A station at which the greatest volume of repairs is expected is the logical first choice. However, other consid- erations may be the frequency with which the operating schedule pro- vides ov.-. .ight layovers, the relative ease of routing other aircraft to that station, the availability of mechanics and parts to support other types of aircraft, the planned volume of scheduled-maintenance work, and so on. Line-maintenance stations themselves vary in size ~ n comp:exity. d The facilities needed for immediate corrective work establish the mini- mum resources at any given maintenance station, but operating organi- zations generally consolidate the bulk of the deferrable work at a few of these stations for greater economy. To simplify the control of scheduled maintenance, individual tasks arc grouped into a fairly small number of maintenanc- ackages for execution. Like deferrable corrective work, these sched tnainte~lancepackages can be arsigned to any con- venient mainr ice station. Thus the more involved work is generally, assigned to th! line stations already equipped with the staff and inventories for ,~,ensivecorrective work. Not all scheduled-maintenance tasks can be carried out at line sta- tions. Major structur-1 inspections, scheduled rework, and inspections which entail exter-s; -+ dissassembly are best handled at a major main- tenance base equipned with shop facilities. The major base also repairs failed units that a. ,moved from aircraft at the line stations. Few such maintenance bases are needed, and reliability considerations generally determine their size and manpower requirements, rather than their location. Many large airlines operate efficiently with only one main- tenance base. The work performed at a maintenance base is generally
  34. 34. termed shop maintenance to differentiate it from line maintenance, which consists primarily of replacing failed units rather than repairing them. The entire process by which a detailed support plan is developed is beyond the scope of this volume. Suffice it to say that a detailed plan is necessary in order to implement a scheduled-maintenance program. Our concern here is with the development of such a program-or rather, L with the principles underlying its development. In the following chap- ters we will examine the nature of failures, the basis on which their con- sequences are evaluated, and the specific criteria that determine the applicability and effectiveness of a given type of prever.live task. With this framework established, we will consider the decisior, 9gic that results in a scheduled-maintenance program based on the actual reli- ability characteristics of the equipment. This reliability-centered ap- proach ensures that the inherent safety and operating capability of the equipment will be realized at the minimum cost, given the information available at any time. The chapters in Part Two illustrate the application of RCM decision logic to specific hardware examples and discuss some of the informa- tion processes i~~volved the continuing evolution of the maintenance in program after the equipment enters service. All these illustrations are drawn from commercial-aircraft applications. However, it shou!d be clear from the discussion in Part One that the basic principles of RCM programs extend not just to otver operating contexts, but to maintenance programs for any complex equipment. ..I i14 INTRODUCTION
  35. 35. CHAPTER TWO THE PARTS of any mechanical equipment are subject to wear, corrosion, and fatigue which inevitably result in some deviation from the condi- tions that existed when the equipment was new. Ultimately the devia- tion will become great enough that the equipment, or some item on it, no longer meets the required performance standards- that is, it fails. The role of scheduled maintenance is to cope with the failure process. For years, h~wever, chief focus has been on anticipating the age at the which things were likely to fail, rather than on how they fail and the consequences of such failures. As a result, there has been insufficient attention to the failure process itself, and even less attention to the question of precisely what constitutes a failure. One reason for this lack of attention has been the common assump- tion that all equipment "wears out" and inevitably becomes less reli- able with increasing operating age. This assumption led to the conclu- sion that the overall failure rate of an item will always be reduced b y an age limit which precludes operation at ages where the likelihood of failure is greater. In accordance with this hard-time policy, all units were taken out of service when they reached a specified age and were sent tc the major maintenance base for complete disasseinbly and overhaul, a procedure intended to restore each part to its original condibion. It is now known that the reliability of most complex items does not vary directly with operating age, at least not in such a way as to make hard-time overhaul a useful concept. Procedures directed at obtaining some precise evidence that a failure is imminent are frequently a far superior weapon against failure. However, to understand the specific nature of such procedures as they pertain to an RCM program, it is necessary to take a closer look at the entire concept of failure. Without16 THEORY AND PRINCIPLES a precise definition of what condition represents a failure, there is no
  36. 36. way either to assess its consequences or to define the physical evidencefor v lich to inspect. The term failure must, in fact, be given a far more , .explicit meaning than "an inability to function" in order to clarify thebasis of reliability-centered maintenance. In this chapter we will examine the probleln of defining failuresand some of t r ~ e implications this has for the analysis of failure data. Wewill also see how failure consequences are evaluated, both in terms ofsingle failures and in terms of multiple failures. Finally, we will discussthe process of failure itself and see why complex items, unlike simpleitems, do not necessarily wear out. 2 1 THE DEFINITION OF FAILUREEach of us has some intuitive notion of what constitutes a failure. We failurewould all agree that an automobile engine, a fuel pumr;or a tire has functional failurefai!ed if it ceases to perform its intended function. But there are times potential failurewhen an item does continue to function, although not at its expectedlevel. An automobile engine may run powerfully and smoothly, but itsoil consumption is high; a fuel pump may pump fuel, but sluggishly;a tire may hold air and support the car, but its bald tread indicates thatit will do neither much longer. Have these items failed? If not, how bad must their condition be-come before we would say a failure has occurred? Moreover, if any ofthese conditions is corrected, the time required for unanticipated re-pairs might force a change in other plans, such as the delay or cancel-lation of a trip. In this event could it still be argued that no failure hadoccurred?
  37. 37. To cover all these eventualities we can define .a failure in broad terms as follows: - A failure is an unsatiefactory cundltion. .- In other words, a failure is any identifiable deviation fmm the original condition which is unsatisfact~ryto a particular user. The determina- tion that a condition is unsatisfactory, however, depends on the con- sequences of failure in a given operating contex;. For example, high oil consumption in an aircraft engine may pose no problem on short- or medium-range flights, whereas on long-range flights the same rate of consumption would exhaust the oil supply. Similarly, engine- instrument malfunctions that would not disrupt operations on multi- engine equipment would be clearly unsatisfactory on a single-engine plane, and performance that is acceptable in a land-based environment might not be good enough for carrier operation. In short, the eract dividing line between satisfactory and unsatis- factory conditions will depend not only on the function of the item in question, but on the nature of the equipment in which it is installed and the operating context in which that equipment is used. The deter- mination will therefore vary from one operating organization to another. Within a given organization, however, it is essential that the boundaries between satisfactory and unsatisfactory conditions be defined for each item in clear and unmistakable terms. FUNCTIONAL FAILURE The judgment that a condition is unsatisfactory implies that there must be some condition or performance standard on which this judgment can be based. As we have seen.. however, an u2satisfactory condition can range from the complete inability of an item to perform its intended function to some physical evidence that it will soon be unable to do so. For maintenance purposes, therefore, we must classify failures further as either functional failures or potential failures: A functional failure is the inability of an item (or the equipment con- taining it) to meet a specified performance standard. A complete loss of function is clearly a functional failure. Note, how- ever, that a functional failure also includes the inability of an item to function at the level of performance that has been specified as satisfac- tory. This definition thus provides us with an identifiable and measur-18 THEORY AND PRINCIPLES abie condition, a basis for identifying functional failures.

×