Quality Assurance & Reliability Engineering PDF

CHAPTERIX QUALITY ASSURANCE & RELIABILITY ENGINEERING 10900 General It shall be endeavour of every official to take steps towards quality in their set-up to improve productivity. Quality Assurance involves effects towards quality improvement, quality development and quality maintenance to meet se...

CHAPTERIX QUALITY ASSURANCE & RELIABILITY ENGINEERING 10900 General It shall be endeavour of every official to take steps towards quality in their set-up to improve productivity. Quality Assurance involves effects towards quality improvement, quality development and quality maintenance to meet service requirements at economical levels. This would require enhancing quality of products, services and activities. 10901 Definition of Quality Use of quality spares for maintaining traction assets plays a vital role in their reliable operation. Quality of a product/ equipment is defined as compliance with the following : a) The specifications as established by the purchaser and accepted by the supplier; b) The design details as declared by the supplier and accepted by the purchaser; and c) Sound engineering practice though not specifically defined either in the specifications or in the designs. A note on scope of quality assurance, quality aspects in manufacturing, system of acceptance sampling, quality, indices for acceptance is given in the Annexure - 9.1 for reference. 10902 Purchase of Spares The following guide-lines have been laid down for purchase of quality spares: (ref. Board's letters nos. 73/RS(G)/ 30/RLL dt. 30.3. 87 & 17.2.89). The various components, sub-assemblies and spare parts shall be purchased from original/approved suppliers. Railways will however make out a compendium of RDSO approved manufacturer's list. Any variation from the same shall only be permitted personally by the Chief Electrical Engineer. 10903 Application of Reliability Engineering To improve quality of service and improve availability of equipment for operations, application of concept of reliability engineering is also being considered as one of the scientific approaches in use. A note on Reliability Engineering is enclosed as Annexure-9.2 for reference. 10904 Employees Participation Participation of employees in Quality Circles and quality improvement is essential for Quality Management of services. It is to be remembered that Quality involves every one in the organisation, management, workers towards improving performance at every level to build an organisational culture where the quality improvement are embedded into the work and the activities. INDIAN RAILWAYS AC TRACTION MANUAL - VOLUME 1  Annexure 9.1 NOTE ON QUALITY MANAGEMENT 1. Scope of Quality Assurance 1) Quality Assurance consists of the measures taken to ensure that three conditions listed below are fulfilled. a) The specifications as established by the purchaser and accepted by the supplier ; b) The design details as declared by the supplier and accepted by the purchaser; and c) Sound engineering practice though not specifically defined either in the specifications or in the designs. 2. The main aim or objective of Quality Assurance is to prevent any defect form appearing or developing in the work done and not merely to detect and reject defective work. 3. While occasional rejection requiring rework or replacement is not ruled out, the objective is to take every possible step to eliminate the basic or root causes of defects. 4. It is also the purpose of Quality Assurance to maintain records in such detail and manner as to facilitate investigations into problems or failures that may arise during the life time of the work done. 2. Quality Aspects in Manufacturing The specific quality aspect in manufacturing includes : 1. Choice of machines, processes and tools capable of maintaining the tolerances. 2. Choice of instrument of an accuracy adequate to control the processes. 3. Planning the flow of manufacturing information and criteria. 4. Planning of process quality controls. 5. Selection and training of production personnel. 6. Planning the quality aspects of purchasing and shipping. 3. Planning Through Trial Lots The trial lot is used to " clear the track" for full scale production by: - 1. Proving that the tools and processes can indeed turn the product out successfully. 2. Proving, on test, that the product will process the essential functional features. 3. Proving, on use, that the product will achieve the desired field performance. 4. Remedying the deficiencies in manufacturing process of product before embarking on full scale production. These proofs and remedies cannot be provided from the record of samples made in the pilot plant. In the pilot plant the basic purpose is to prove engineering feasibility, in the production shop the purpose is to meet standard of quality, cost and delivery. The pilot plant machinery, tools, personnel, supervisions, motivation, etc. are all different from the corresponding situation in the production shop. INDIAN RAILWAYS - AC TRACTION MANUAL - VOLUME I 4. Acceptance Sampling 1. Introduction 1. 1 Acceptance Sampling is the process by which decisions are taken either to accept or to reject an entire 'Lot' of products offered for inspection, on the basis of detailed 100 percent inspection of one or more samples drawn at random from the lot. 1.2 The number of items to be drawn from each sample, the number of samples to be drawn from the lot and the number of permissible detectives in each sample, constitute what is known as the Sampling Plan. 1.3 Acceptance Sampling is based on the mathematics of Probability and Statistics. Sampling plans are generally selected from published tables to suit the expected quality levels. 1.4 The following Indian Standards must be studied by all Engineers concerned with Inspection and Quality Control. a) IS 397 -Methods of Statistical Quality Control During Production b) IS 1548 -Manual on Basic Principles of Lot Sampling c) IS 2500 -Sampling Inspection tables Part I - Inspection by Attributes and by Count of Defects. Part II - Inspection by Variables for Percent Defective. d) IS 5002 - Methods for Determination of Sample Size to Estimate the Average Quality of a Lot or Process. 2. Quality Indices for Acceptance Sampling 2.1 Acceptable Quality Level (AQL): This is usually defined as the worst quality level that is still considered satisfactory. The units of quality level can be selected to meet the particular needs of a product. Thus, “MIL-STD-105 D" defines AQL as "the maximum percent defective (or the maximum number of defects per hundred units) that, for purposes of sampling inspection, can be considered satisfactory as a process average." If a unit of product can have a number of different defects of varying seriousness, then demerits can be assigned to each type of defect and product quality measured in terms of demerits. As an AQL is an acceptable level, the probability of acceptance for an AQL lot should be high (see Figure 9.01) 2.2 Rejectable Quality Level (RQL): This is a definition of unsatisfactory quality. Different titles are sometimes used to denote an RQL for example, in the Dodge-Romig plans, the term "Lot tolerance percent defective (LTPD)" is used. As an RQL is an unacceptable level, the probability of acceptance for an RQL lot should below (see Figure 9.01). In some tables, this probability is known as the consumer's risk designated as Pc and has been standardized at 0.1. The consumer's risk is not the probability that the consumer will actually receive product at the RQL. The consumer will in fact not receive 1 lot in 10 at RQL fraction defective. What the consumer actually gets depends on actual quality in the lots before inspection, and on the probability of acceptance. INDIAN RAILWAYS -AC TRACTION MANUAL VOLUME I 2.3 Indifference Quality Level (SQL): This is a quality level somewhere between the AQL and RQL. It is frequently defined as the quality level having a probability of acceptance of 0.50 for a given sampling plan (see Figure 9.01). 2.4 Average Outgoing Quality Limit A relationship exists between the fraction of defectives in the material before inspection (incoming quality p) and the fraction of defectives remaining after inspection (outgoing quality AOQ) : AOQ= pPa. Obviously, when incoming quality is perfect, outgoing quality must likewise be perfect. However when incoming quality is very bad, outgoing quality will still be perfect because the sampling plan will cause all lots to be rejected and detailed inspected. Thus at either extreme-incoming quality very good or very bad- the outgoing quality will tend to be very good. Between these extremes is the point at which the percent of defectives in the outgoing material will reach its maximum. This point is known as the average outgoing quality limit (AOQL). 3. Sampling Plans 3.1Normally, Sampling Plans have to be specified clearly by the purchaser because it is essential to have agreement on this issue between the Purchaser and the Supplier. When so specified, the sampling plans should be followed scrupulously by the Inspector. 3.2When the specification does not include Acceptance by Sampling and if the Inspector considers that 100 percent inspection is neither practicable nor necessary, a reference should be made to the purchaser. 3.3Different Sampling plans may be adopted for different properties or parameters of the items to beinspected. More important properties may be checked on sample of larger size. INDIAN RAILWAYS — AC TRACTION MANUAL - VOLUME 1 3.4 Where the items being inspected are being produced by fully automatic machines the sample size specified may be smaller than in the case of manually produced items. It is desirable in such cases and also in general when lots of identical items are inspected repeatedly, to carry out the inspection in the same sequence as manufacture. Control charts should also be maintained to enable timely detection of dimensions going beyond permissible limits. Whenever the machine is reset or retooled, 100 percent inspection should be carried out by the manufactur's Inspection organisation until consistently good results are obtained. 4. Random Sampling The conclusions drawn, relating to the quality of a whole lot on the basis of a 100 percent check on a sample, can be relied upon only if the sample is sufficiently large and the sample is selected in a totally unbiased or random manner. Above all, the selection should not be influenced in any manner by the Supplier's suggestions or actions. The quality of acceptability of an item should not determine the choice of items for the sample. Even the inspector himself should try to avoid any inadvertent or unintended bias in the selection of the sample. This can be ensured by using a table of random numbers. 5. Sample Size 5. 1 One item drawn from a lot of any size cannot be relied upon to give any idea of the quality of the lot except in the case of fluids or fluid-like fine powders which have been thoroughly mixed together. In these exceptional cases a small quantity-apparently a single sample-will give a reliable measure of the whole lot. 5.2 The Sample size, i.e. the number of items included in the sample, is very important. It increases with increasing lot size-but not in the same proporation. For example, for an AQL of 0.65 :- Lot Size Sample Size Sample Percentage 100 20 20 1000 125 12.5 3000 200 6.7 6. Limitations of Acceptance Sampling 6. 1 The limitations of Acceptance Sampling are as follows : a) Acceptance Sampling involves some uncertainty or risk. For instance, it is not possible to say with certainty whether any individual item taken from a lot which has been accepted by Sampling, is good or bad. Therefore, Acceptance Sampling should not be adopted where defects are not acceptable in even one item as, for example, in Safety Items. b) It is possible to estimate the percentage of good or bad items in a lot accepted by sampling with adequate accuracy and confidence, only if all the rules of sampling are followed strictly. c) If an incorrect sampling plan is adopted or if sample selection is not properly randomized there is danger of very bad lots gettng accepted. 6.2 Where large quantities or numbers of identical items have to be inspected, the advantage of acceptance Sampling outweigh its limitations. Moreover, destructive tests have necessarily to be based on 'Sampling' basis. INDIAN RAILWAYS --AC TRACTION MANUAL VOLUME I  Annexure 9.2 NOTE ON RELIABILITY ENGINEERING 1. Introduction Reliability Engineering is not totally new to engineers. They have always been practicing the essentials of reliability engineering without giving it this name or using its terminology. 1.1. Reliability engineering is the discipline concerned with the prevention of defect, failures, fires, accidents etc., in all types of hardware from the smallest items like hand tools to the largest units like loco- motives, boilers, turbines etc. This discipline has been successfully applied in a number of complicated and baffling cases to reduce the breakdown rate of equipments, improve the availability of plants for operation and thus to help in reduction of costs and improvement of efficiency and productivity. 1.2. Even though the overall or general design of the multitudes of types of hardware are widely different, the detailed design of components as also the mechanisms of failures are generally similar. The basic principles of reliability engineering can be applied to identify the root causes of failures such as weak links in the systems, starting points of material failures, causes of workmanship defects, degradation processes, and many other such factors which usually lie hidden under the obvious causes of failures. 1.3. Reliability engineering is most appropriate for repetitive types of failures which continue to occur again and again despite various measures being taken to avoid these failures. In such intractable cases, the user has either to suffer the failures continuously or to take the burden of replacing the entire equipment in question. Such intractable cases have been solved economically by the application of the principles and methods of reliability engineering. While the basic principles of design, manufacture and maintenance differ widely between the fields of civil, electrical, mechanical or signal & telecommunication engineering, across all these specialist branches cuts the new disciplines of reliability engineering. The modes and mechanisms of failures of all types of hardware are the same and the statistical/ mathematical methods for understanding and studying them all, are identical. 2. Basic Principles 2.1Definition of Reliability Formally, reliability is the probability that an item will perform as required, under stated condition, for a stated period of time. Thus if we have a large number items on test, we can write:- Number surviving at present Reliability at time t = R(t) --------------------------------------- Number at start 2.2.Failures When an item no longer works as intended we say it has failed. Therefore, "Failure is the termination of the ability of an item to perform its required function. 2.3. Classification of failures. Failures can be classified as follows : a) As to Cause: 'Misuse failure' is a 'failure attributable' to the application of stress beyond the stated capability of the item. Thus it has been ill treated. An 'inherent weakness failure' is a failure inherent in the item itself, INDIAN RAILWAYS - AC TRACTION MANUAL - VOLUME 1 b) Constant failure period: Once the early failures have been removed, the parts usually settle down to what may be a relatively long period, when the failure rate is approximately constant. During the constant failure period, it is usual for failures from a wide variety of causes occur at random with no obvious pattern, except that the failure rate is roughly constant. Such failures are also commonly called random or chance failure. Where failures do form a well defined pattern the reliability engineer calls them 'systematic failures' and such patterns usually provide valuable information about the cause of a failure. c) Wear out failure period: The incidence of failure in this period is high since most of the component will have exceeded their service life and consequently would have deteriorated. 2.5. Failure Mechanism: A few of the common failure mechanisms are Adherence Deterioration Piezo electric defect Arcing Diffusion Radiation damage Backlash Drift and shift Secondary Current Bleeding Dynamics out of limit Seizure Carbonisation Electric Breakdown Silver migration Composite behaviour Erosion Slip Contact bounce Fatigue Smearing Contamination Fretting or galling Sublimation Corona Frequency effect Voltage Breakdown Creep Leakage Voltage overload Creep rupture Magnetic hystreresis Wear Cross talk Mass unbalance - Current Overload Noise Decarbonisation Opens 2.6. Mean Time Between Failures (MTBF): Where the failure rate is approximately constant, it is convenient to use the Mean Time between Failures (MTBF). The mean time between failures is the reciprocal of the failure rate. It should be as long as possible. 2.7.Mean Time To Failure (MTTF): The term mean time to failure is analogous in every way with mean time between failure and is used where a failure can not be repaired. INDIAN RAILWAYS - AC TRACTION MANUAL - VOLUME I 2. 8. Mean Time To Repair (MTTR): It is the mean time taken to put the equipment right after it has failed. It should be as short as possible. 2.9. Availability: It is the probability that an equipment will be available for use and is given by the following relationship: MTBF Availability = ----------------- MTBF + MTTR 3. Failure mode effect and criticality analysis (F-MECA) a) F-MECA method is used very widely at the design stage for estimating the reliability of any new system.or product and, more importantly, for evolving more reliable designs through the identification of vulnerable or failure prone components. b) The method is described in United States MIL - STD-1629 (Procedures for Performing a Failure Mode, Effects and Criticality Analysis). The basic principle of this method consists of listing all the components of an equipment or system and to evaluate, the effects of each possible failure mode of each component on the equipment or system as a whole. The results of failures are classified according to the severity of the effects. c) This method was originally developed mainly for electronic equipment but it can be applied equally well to heavy electrical equipment or even mechanical equipment or systems. d) Formal application of F - MECA methods is certainly useful for evaluating or designing complex systems but there is another advantage of learning this system. It gives an insight to the engineer which is useful for not only designing but also for investigating failures of small equipment or components. Therefore, it is useful to study this system for even those who will not be called upon to apply it for designing, evaluating or improving complex systems. e) Although F-MECA methods as defined in MIL-STD-1629 were originally developed for the purpose of evaluation of reliability of electronic equipment at the design stage, it is possible to devise a variation which is suitable for a complex operational system such as the Railways, not so much for designing the system as for evaluating the effects of various failure modes on performance. Such an analysis will help to place in the proper perspective different types of failures which occur everyday. The overall picture -so produced will help the top management firstly to determine the most effective application of available resources and secondly to asses the effectiveness of measures taken to prevent any type failure. 4. Failure Reporting and Corrective Action Systems (FRACAS) a) FRACAS system was originally designed for reporting and investigating systematically all failures which occur during the development of any new product. However, it can equally well be applied to an organization such as the Indian Railways to systematize the action that needs to be taken to improve the reliability of the hardware. It is as important for old designs of equipment which have been in service for many years as for new equipments introduced into service recently. If the very first failure of each type on any new equipment is treated as a problem to be investigated and corrected much time and money can be saved. b) The starting point for the establishment of a FRACAS is the constitution of a failure review group (FRG) comprising: INDIAN RAILWAYS - AC TRACTION MANUAL - VOLUME Divisional officer in-charge of maintenance. HQ officer in-charge of design/development Any other officers who could contribute to the solution of the problem. c) The FRG should work as a team to solve the problem and not as a forum for fixing responsibility, preparing a report or commenting on failure reports. Its recommendations should aim at practical and economical solutions. Where action is within their own powers or capacity the FRG should implement their decisions, if not, they should consider it their responsibility to obtain the required approvals from higher levels. They should always keep in mind that the effort must always be to determine the technical solution in detail. Merely reporting the problems to manufacturers for necessary action is not the function of an FRG. d) The FRG should meet at least once a week to review the failure reports on the equipment under consideration and to determine the corrective actions. e) Whenever any new equipment is commissioned an FRG should be set up. Similarly an FRG may be established for a few specific problems of old standing which may be causing concern. The FRG should remain in force until the reliability of the equipment attains the desired level. f) The starting points for an FRG are the source reports on defects, failures, observations on operating irregularities etc. United States MIL-STD-781 provides a description of failure reporting methods. The common elements of all such failure reports are: - description of failure symptoms - effect of failure - immediate action taken - elapsed time after commissioning, after various maintenance schedules. - operating conditions - date, time, place of failure - make, type, serial number of equipment and of component(s) - opinion of person writing the report regarding possible cause of failure - design modifications, if any. g) Failure report forms should be designed to cover all the above details as also any other relevant data specific to the equipment in question. h) Corrective action proposed by the FRG and approved for implementation should be clearly defined and its implementation on the entire population of the equipments in question should be watched. Then performance after the corrective action should be monitored. The FRG for any particular equipment may be wound up only when the desired reliability level is attained. 5. Potential of Reliability Engineering 5. 1 Design Stage i) If the specification is drawn up carefully, it takes care of what we may call the 'Gross' design requirements. If the equipment complies with the specification and passes the tests stipulated therein, it will meet all the performance requirements and is unlikely to have any major defects which could render it unserviceable in a short time. However, this is not enough. The number of ways in which things can go wrong is so large that careful scrutiny of design details in every component is essential. Such scrutiny is inescapable in the case of equipments being manufactured by new firms for the first time. Even in the case of products made by reputed firms, if it is a new product, it is desirable to carry out such detailed scrutiny of designs. Very often, apparently minor deviations or discrepancies. INDIAN RAILWAYS – AC TRACTIAON MANUAL VOLUME I can lead to avoidable failures in service. It is in the interest of not only the user but also the supplier that the detailed working drawings are scrutinized by reliability engineers familiar with the performance of the equipments under actual service conditions. The only time such designs scrutiny can be waived is when an identical design has been in successful service for many years under identical conditions. Even in such cases it is necessary to check that there has been no change in the detailed design of any component. Further, the opportunity should be taken to review the reliability and failure statistics to see whether any improvements in the design of the concerned components can be introduced. 5. 2 Manufacturing Stage i) It is always not possible to visualize every thing at the drawing board stage. When the manufacture of the high prototype is undertaken, various problems may be exposed, particularly in the case of complex systems. The problems may relate to manufacture or to maintainability and reliability. Studies relating to maintainability/reliability must continue concurrently during the manufacture of prototypes. ii) As far as the actual method is concerned the best way in the long run to ensure reliability is to insist on a fanatic or rigid adherence to drawings, process sheets and such other production documents. Relaxations/deviations and even so called 'improvements' in process should be reviewed carefully by both basic design engineers and reliability engineers before permitting them on the production line. If such scrutiny reveals an unnecessary or superfluous feature in the drawing, the drawings or processes should be modified but as regards the production staff are concerned, production documents mentioned above viz. working drawings, process sheets, etc. should be treated as sacrosanct. 5.3 Inspection and Testing i) While the function of inspection and testing organization is to carry out the actual inspection and testing, it is the function of reliability engineering service to define what these inspections and tests should be, when they should be carried out and what criteria should be followed and so on. ii) Organization of stage inspection from raw material stage through components and sub-assemblies upto the final inspection and testing of the finished product is of the utmost importance. There are many types of defects in regard to tolerances, material specification and process parameters which will have little or no effect on the performance of the equipment as may be judged during acceptance tests or even limited actual service but such defects can cause even catastrophic failures in actual service. These failures may occur at any time. Some may occur within days after commissioning whereas some others may develop after several years service. It is the function of the reliability engineer to investigate all such failures to determine the real or probable causes and to alert the stage inspection organization to watch for and eliminate these types of defects. iii) In this connection, special mention must be made of screening or burning-in procedures. Many components, particularly those which require sophisticated technology in production or those in which contamination and invisible dimensional inaccuracies can cause failures, exhibit a high rate of failure, initially. This is termed as 'Infant mortality'. Examples of components which exhibit this pattern of failures are semi- conductors, fuses, incandescent lamps, coils with very fine wire etc. Reliability of such components in service can be improved by operating them at a stress level which is significantly higher than that in service. Weak components which would otherwise have failed in service will fail or at least show some deterioration of properties during the screening procedures. By eliminating such components the reliability of the remaining components which survive the screening process will be much higher in actual service. It must be noted that 'screening or burning - in' does not improve the quality of the components. It merely accelerates the failures of those which would have failed any way in service. 5.4 Operation and Maintenance i) The most important contribution that can be made by operating staff and maintenance staff towards the improvement of reliability is in regard to investigation of failures. It is necessary to determine and INDIAN RAILWAYS AC TRACTION MANUALVOLUME 1 analyze as accurately as possible the actual conditions of service under which the failures occur and to re-construct from the observed data and study of failed components the exact mechanism of failures. On the success of these studies depends the evolution of the quickest and most cost- effective solutions to the problems. Experience in failure investigations is invaluable for this purpose but a systematic study and application of the principles of reliability engineering can greatly speed up the investigation. A through insight into the design of the equipment with regard to the calculation of various types of stresses is of-course desirable and often essential but it is also possible to evolve elegant solutions without going into the full design details. 6. Reliability and Cost a) It is not always necessary that improvement in reliability must cost more. Better designs, different materials, better quality control during production etc. may achieve improved reliability at little or no extra cost. If scrap is reduced at the same time, the overall cost may actually come down. Predicting the cost of achieving any given reliability is nearly always difficult, so that in general we only know the cost accurately afterwards. b) Maintenance costs are also difficult to estimate. If we know precisely what repair work has to be done each time an equipment fails, then we can predict its cost. However we are unlikely to be able to foresee associated costs such as:- i) The value of production/service lost through breakdowns. ii) The cost of having the equipment of action. 7 Suggested Books for Further Reading 1. Practical Reliability Engineering by Patric O'Connor (Published by John Wiley) 1. A Practical Approach to Reliability by R. H. Caplen (Published by Business Books Ltd. *3.MIL-HDBK-217C: Reliability Prediction for Electronic Systems *4.MIL-STD-1629: Failure Mode, Effects and Criticality Analysis *,5 MIL- HDBK-189: Reliability Growth Management. *6.MIL-STD-781 Reliability Qualification and Production Approval Tests. 7- Selection and use of Engineering Materials by FAA Crane and J.A. Charles (Published by Butterworths). * United States Military Standards obtainable from National Technical Information Service, Springfield Virginia, USA. INDIAN RAILWAYS- AC TRACTION MANUAL VOLUME I