— Reliability DEFINITION Reliability is the probability that a device tory without failure, for the stated period of ied operating condit In the above definition, there are 4 factors which are essential to the concept of (0) Operating environment should be defined. t Bath ub curvetaue rte curve fare ato > Zone 1 —Early Failure Period It Js characterised by a high initial failure rate, failure rate. By the end of the failures in this zone are assignable causes 1. Design deficiency 2. Manufacturing error 3. Raw material defects 4, Wrong maintenance practices. dropping off to a low ne or more of the foll (SueuT 207 tone 2 — Random Failure Period characterised by a more or less constant failure rate. This ts the rate at which usage of the product occurs without any expectation of failures. 1res in this zone are due to chance causes. tone 3 — Wear Out Failure Period It is characterised by a gradual increase in failure rate. Failures in this zone are due to one or more of the foll 1, Ageing tion in physical strength properties Relationship between the failure rate and the mean time between failures (MTBF), |, MTTF (Mean time to failure) efined as the mean time interval between succes he MTBF is the failure rate, which is denoted by where n= No. of items falled, T= tested duration. = Prove that the Failure Rate is the Reciprocal of the MTBF Proof, Let us test ‘items each thours and the items which fail are repairable. ise there are 'r failures, 26 Items failed _ "Total test duration nt The failure rate, = (0) rents empp = = Teal testduration _ nt = > x96 of items falled ~ 1 from (1)‘Solution. Data : MIBF = for t= for t= 10 hr, R=0,9048 for t= 50 hr, R= 00180) R= 0.6065. for t= 100 hhr, R= 01100 R= 0.3678 for t= 200 hr, R= e100! *200 R= 0.1853 for t= 300 hr, R= e201 *:00) R= 0.04978. ts useful Ree 610-6100) R= 0.9995 2 (© wer, a F= 10,000 x 100 x 5 x 10-8 = ra5 ‘ (9 MTBF = (@ Ree Systems Reliability Re 810-6 2.00,000) System connected in series R= 0.3678, System connected in parall () Ree ‘System connected in series R= @ 5 10-. 1,00,000) R= 0.6065. Ifa system consists of 3 components A, B, C in series, then the reliability of the ‘system, 3. A piece ground support equlvlent What ists reliably for amlssion te of | hs 10 Graph these ansiwers by plating Rg = RyRy - Re ~ System Connected in Parallel Here the function of A can be done by B or vice versa. If the system consists © 's Aand B in parallel with reliability R, and Re, then the reliability of the system component Benicia[zo REDABILITY TOTAL QUALITY MANAGEMENT © R= Ree = 0.9999, ‘components are connected in parallel, Re=[1-(1- Rg - RJA -Rol nents of the syaten SM Fate ofthe system = the sum ofthe 619 = 0.9998) or 6.10 equal PT. for a series system: the falure rates are acl wie aee Le PT gm hg thy +n + = Rihen-oom) Proof. Let Rs, Ry. R rn bu Ry snes Be Felablity of the systen ates ty system and its com, Ra 0.200 For independent compor 2. Aseries system has 3 independent parts A, B, C which have a MTBF ef 100, 400, 4n series and exponential Ro= Ry. Ry Re 800 hrs. reliability. Whigs gy Ags Pag vere are the fal Find (a) MTBF of system. mean time ‘.. ‘he failure rate ofthe system and its compone (b) Failwe rate of system in failures /milion hrs. From eqn, sailure rate of system in percent failures/ 1000 hrs. (d) Reliability of system for 30 hrs. ‘The increase in MTBF of A component to a.30% of inerease in MTBF of system. Solution. Data : MTBF, = 100 hrs. ; MTBF, = 400 hrs. : MTBF = 800 hrs. Pa (em Pan) Beets nas iy = 0.0 ate Since the bases are save, bj 4 x tos tata igre sa Gig = 0.0005 atc Hence proved, 1. A system has 3 uni = 0,00125 failures hr. eae ‘hg = 0.01 + 0.0025 + 0.00125 2g = 0.01375 fallures/hr, 1 a Aya 4% 10° failures/hr. : he = 3.8 x 10°° fat | ee eee hc = 3.8 X 10° failures /hr. = (@ Assuming that components are in series, eee () WET, 25 = 0.01875 fatlures /hr. Fat hy the 5 = % 10,00,000 failures /milion I. (0. milion = 10%) % wry, 2 = o7sna ne | 7a 0olare i aneare aay 5 2.9 18750 fallures/million hr. Boa(lb1 4586) x10°° {9g = 0.01375 x 100 x 1000% failures 1000 hrs. 4g = 9.8 10° fitures/hr Vig fash | (60) (1000 hrs.) fatlure rates can be added and not the MTBFs, To find MrBrs= 2 | 4g 1875 percent fatlures/1000 hra.3. A system ts composed of 10,000 parts. What averag Achieved to get a system's MTBF of 25 Solution. Data required ? Solution. Data : MTBF = 50 hrs ; t= 45 R=e% TOTAL QUALITY MANAC ge. aa] [Reuasiry 0.7176 New = LATS «0.0159 ee ages O 1 = —1_ -6289 Trew” 0.0159 = 25.78%, 50 hrs. Calculate the reliability for 75 hrs of operating ‘creased by 10%, 20%, 30%, 409, 50%, caleulate (a) the seccssary (2) plot a graph % change in reliability v/s (TBF = 50 hrs ; t= 75 hrs. a 1 = pgp gp 7 0-02 alures/r Reet = e279 20.9931 (@ @UR=11x te, 0.2454 = € in0.2454 =~ (x75) = 1.4048 =-2.* 75 = 0.0187 failures/hr, => MTBF = 53.48 hrs. (FR = 1.2 x 0.22% te, 0.2677 = € in0.2677 =- (x75) = 1.3178 =-2.x75 = 0.0176 fatlures/hr. > MTBF = 56.82 hrs. = 60.61 brs, 0.8123 ifR= 140.2281 te, 0.9123=¢0*7) In0,3123 =~ 0.75) =1.1687 =-2x 75 = 0.0155 failures/hr. => MTBF = 64.52 hrs. similarly sf R= 1.5 x 0.2231 = 0.9347 te, 0.9347 = ¢0*7) In 0.9347 =- (2.x 75) = 1,0947 =— 2.x 75 4 =0.0146 failures/hr. = MTBF = 68.49 hrs.TOTAL rY MANAGI > MTBF (9 » 5-53.48 100 50 =~ 6.9696 % change in MTBF (ig = 50=56.82x100 __, 3 = 136 0 = 60.61 100 cee == 21.22% 9% change in MTBF (iy) = 5-64.52 100 30 =~ 29.04% % change in MTBF (x) = 50=68.49%100 __ , =42*100 « _ 36.00% Fee Ace sinter sorts pe rate ts 20 failures/10° hrs in series. The e ty. Solution. 2, =30 x 10°® failures/hr, 2, = 25 x 10° failures /hr Jog = 20 x 10°® failures/hr. Since the components are connected in series, R= 0.9851 7. A cassette player has got 4 subsystems nam Pick up head, amy formance of a cassette player, all various subsystems are speed control system, the magnetic ier systems. For a satisfactory per- e 4 systems must perform satisfactorily, The MTBF of th u See Er Cl His vehicle can be imagine 715} = 4000 hrs. yer per 1500 hrs reception 14x 10° failures/hr. Aur = 3560 haanas = Gag 7 225% 10°* failures /hr. bases ~ 000 1 6 hos = qos = 9:983 « 10° fallures/b. os * 3000 1 the components are 215x104) +4 x10) connected in series, (2.5 x 10°4) + (3.33 x 104 1 ___1__ _.675.67 hrs. 48x10 re rate not >r than 0.1% per 1000 hrs of operation. we rate, what {s the probability that one of these units 2000 hirs of service ? le failure rate where the probability of survival for & 1 = 0.1% per 1000 hrs. 3 100 hrs. = 0.1 x 10°2.x 10°® failures/hr- la Re = 1x10" 52000) = 0,098 wo 5 x 107 failures/hr. eee ly ts required to complete a ono so heoe 3 subsystems namely fuel ignition and OOS 9. A participant in a motorBoas The mean tne to repair MTTR) of the 3 subsystems are known 6000, 8000, 10000 kms respectively. Find the relic ‘id ston wothout repair Solution. MTTR, = 6000 kans MTTR, = 8000 kms MITRo = 10000 kms, Dea re 2" Farm, ~ Gang = -66%10" fares ins, ) y= GA = 1.25 x 107 fattures/m. Important 1 4o™ T5999 = 1x 10° fatlures/km, fe connected in series, 1 x 10 failures /km, 92300) 10. What ts the faure rate of a plece of equipment {the prob, of survival for $00 of eperaing period PEpres te fara en She Solution, Data. R = 0.88; t= 900 hrs; i=? R=e% 0.88 = 0900) -0.1278 =-2x 900 A= 1.42 x 104 fe = 1.42 x 104 i= 14.2% failures/1000 hrs, 411. The MTBF of an equipment ts 1000 hrs. What Jailures/1000 hrs. (2) failures per 10° hrs. (0 Saiture-free period, 1 1 ton. 2= 1 _- _1__ igs Solution. 2 3 10° fallures /hr. te expressed tn (a) % ‘MTBF a guaranteed depends upon) A= 10 x 100 x 1000% failures/1000 hrs, 12. A750 hws. life test is performed ef operation. All others survive the ‘components. I component falls after 350 hrs. test. Compute the flue rate. No.of items fated 1 WET, 2 = No.of items fled _ poe. Total test duration ~ & = 2.44 x 10" fares ir.” lability ef his completing the me. | REDUNDANCY j (1 -Rg) = Unreliability of the system. a {fr components are connected in parallel with the same reliability ; ses connective) ‘compared to n inereases the reliability when. is by utlising the Carll omen meee he ey he aren eat ar ‘edunancy, Tenant relay oe eS tm one compre em to perform te same fonction. In #4 fystem to er ae the ester fare, since addtional COmponcsis fanction. to perform the same res dundanoy ts defined asthe characte compe attics are reverted fom causing sysem faies which has the to inerease the reliability ofthe system, select the component ity and then arrange It parallelly. ‘which marginal ictnepeemee dead | Rg = Ry Re) Re Ry = 0.8 x 0.75 x 0.7 = 0.44. a, Jets clear thatthe Rist from 028 490 {ion of Improvement Factor : ‘ ae a prt retunants, te improvanent a 1-R Fei where (1 —R) = Unreliability of each component. belo Rg 21-11 = Rgll -Rp) ne “RYSESE — TOTAL QUALTY MANGER] [aan Ryscot = Rasen - Re (b) How wil the reliability improve further if the | dundart ? Show the ‘fo}— RYO - Ry -RJ}= 1-1 BP ~08) 12 - 0.801 = 0.1202. the foll System with all units operating. P,As=a ieee tee Re = 16.9000» 10-4 x 800) Rs = 0.5757. Solution, gaa Reert Re ©6255 105 +1800) R= 0.9041 cere fac ration gr ae: tab of n= 0.01 Ag = 0.015 2e= 0.02 p= 0.02 4p = 0.025 72500 = 0-00008 faitures /hr. Yat hy the + 2p + Ap = 6.9009 x 107 = 1.56 x 10°® failures hr. = 2x 10° failures/hr = 1,7 x 10°? failures/hr. % 10° failures /hr. TOTAL QUALITY MANAGEMENT tem for 20 hrs. of operating period. The configu: Ryaco * Re Ryneoe = 0.9172 x 0.6085 ment tay Tet} TE oan oo os 085 O85 LHe Solution. Ryiower1 = Rs “Fat Ro- Be = 0.9 x 0.92 x 0.85 x = Ryjovera = 0:6686 Royton * Roamer: Re -By = 0,8902 x 0.99 * 2244 tet reheat the furnace using pulverized cool Trcludes, stator ne blower system Te ents of 0.9, 0.92, 0.85 and sm has a reliability of 0.99 and the ‘ots, Dra die system confoue 0 fr 160 hs. of operator | sa ere ca 0.95 = 0.6686. rnnected in parallel, 0.98 = 0.8872.1.2 FAILURES OF ENGINEERING SYSTEMS The failure of a system or a component is the inability of the system or component to deliver its intended function satisfactorily. Failures may be - either partial or complete. For example, a 2 HP motor has completely failed if it does not produce any power. If it produces some power (say, 1 HP) then the motor has failed partially. Since the term ‘partial failure’ is subjective in nature, it is required that we clearly define a failure in quantitative terms, which can be understood without any further explanation. Some systems have operating ranges. For example, a voltage stabilizer has operational range of 230 + 1% V. Therefore, if it produces output voltage within this range, then it performs well; otherwise it has failed. There are other cases of failures in which the physical failure of a component is considered a success. For example, consider the case of a protective fuse wire. If the fuse wire burns (physically destroyed) when there is overload, we say that it is a success. In other words, we say that a fuse has failed when it does not burn when overloading occurs. There are other devices having multi-model failures. For example, a mechanical valve may experience failure to close, failure to open, or failure to stop leaking. There may also be complete or partial failures. Again, in many other cases, a failure is defined based on the particular application, and not only on the intended design function. For example, consider a voltage stabilizer designed for 230 + 1% V output. If we use this for an application where we require only 230 + 2% volts, and if it supplies this, then it may be considered a success. On the other hand, the same stabilizer is a failure if it does not support an application where 230 + 1% V is required. Therefore, it is essential for any reliability activity to start with a clear and objective definition of failures, preferably in quantitative terms, to avoid any confusion at later stages. This further reinforces the need to clearly specify the intended function and state the definition of failure of a system before attempting to evaluate its reliability. 1.3. CAUSES OF FAILURES ‘There are manyChapter 1: Reliabilty Engineering, = fes of Engineering Components and Systems 1G.No. Causes of failure Details 1 Poor design ‘Wrong material, wrong dimensions and tolerances, improper design models without considering the load and its fluctuation, stress concentration, inadequate interface di lack of process design, no interchange: of parts ‘outdated technology and old machines, 2 Incorrect manufacturing 3) Improper testing Complexity is another reason 4 Complexity number of components and interconnee- more number of interfaces Raw material supply Quality assurance struments. and ipling techniques water transportation requires ing with shock resistance, and environmental protection. Lack of this damages “the system during transportation : Improper foundation, excessive vibration, inadle~ ‘quate inputs (voltage, current, etc), bad quality accessories (switch plug, ee), usage ofVG" eainy Eogheesing and Lie Testing r Z eee uieteng gs 5 eee TABLE 1.1 Causes of Failures of Engineering Components and Systems (Contd) ME TE ae Somme SINo, Causes of failure Details ee ee naa 10 Operational Wrong instruction, lack of clar instruction Understand, poor language of manual 11 Human error Lack of understanding of process and equipment, carelessness, forgetfulness, poor judgmental skills, physical disat tigue Ee 1.4 RELIABILITY AND QUALITY Both quality and reliability are indices used to meast 4 product or system. There are many det the performance of According to one which has passed all qual is the ability of a product to time under stated operating conditions. quality system using poor quality comp to construct a high reliable system redundancies. However, the techn 1.5 REPAIRABLE AND NON-REPAIRABLE SYSTEMS In general, engineering systems can be classified into repairable sy non-tepairable systems. Non-rep: ems are those systems WI s are over as soon as the wre occurs. These cannot be repaired and reused. For example, a a non-repairable system. An electric bulb, a satellite, a non-degradable battery are other examples. These are one-shot devices; we may call them ‘use and throw’ type devices. Repairable systems, on the other hand, are those systems which undergo many cycles of failure and repair within the duration of their de - Most of the industrial machinery and consumer products are repairable systems. The life of such systems does not end with a single failure; it is much longer. As a general tule, reliability of non-repairable systems must be higher than that of repairable systems. For repairable syste . appropriate to use the terms the system performance. These are di Preventive and predi 8 may be reliability of non-repairable systems. Whereas alltvnes of clvneouaes _ aaa ~ Mean Time Between Failure (MTBF): » 5 Chapter 1 : Reliability Engineering * 7. 1.6 RELIABILITY CHARACTERISTICS ‘A number of reliability characteristics are used for studying the performance of engineering systems. Important ones are briefly discussed below: Reliability: Reliability of a product or system is the probability that the product performs adequately its intended function for a stated period of time ‘under stated operating environmental conditions. It is a function of time. It can be shown that the times to failures I distribution. The expected value known as the MTTF. This parameter is ly used in industrial practice compared to reliability. Mean Time To Failure (MTTF): ‘This parameter is similar to MTTF, epairable systems. This is evaluated as the expected value of al distribution of time between failures of a system. od fe ‘Mean Time To Repair (MTTR): It can be shown that the times to repalt of a product or system follow a I distribution. The expected value of this distribution is known as the MTTR. This term is used for repairable systems. Availability: Availabilty of a product or system at any time is the probabillty that the product or system is in the operational condition at that time. ft ix ‘atime function. It can be seen that after a sufficiently long period of operation, lizes and remains constant thereafter, This product or system avail constant value of availability is known as the steady state availability (SSA), It can be shown that SSA over a period of time is the ratio of total up time to the total of up time and down time of a system during the same time period after reaching the steady state. Its also evaluated using the relationship‘ ‘ B © Reliability Engineoring And Life Testing ' ee ne ee nea ee are required for quantitative reliability analysis. For this purpose we require failure and repair data of products and systems. Data can be obtained from various data banks, MIL-HDBK-217F is an excellent source of reliability data. Data can also be generated in-house by means of conducting various types of life tests, Most part of this book is focused on planning and conducting adequate types of life tests for data generation, and various methods of data anal and data management for quantitative assessment of reliability characteristics. 1.7, THE BATHTUB CURVE We have seen that ility is a birth-to-death phenomenon, covering the entire life span of a product or system. It has been observed that if we conduct life tests on any type of product or system, and plot the hazard rate as a function of time, it follows a pattern similar to a bathtub as shown in Figure 1.1. Mission life Operational life Infant mortality Ageing >! Hazard rate Time FIGURE 1.1 Bathtub curve. Due to its special shape, this curve is popularly known as the bathtub curve in the reliability literature. This curve has three distinct parts. Initially large number of failures has been observed resulting in high hazard rate. This Part of the curve is popularly known as the infant mortality period. The number of failures observed during the initial period is mainly due to bad design, wrong manufacturing process, inadequate selection of material and poor quality. The hazard rate then gradually reduces and stabilizes to a constant value, which corresponds to the useful life period. No specific reasons can be assigned for the type of failures that occur during this period, Failures occur purely by unexplainable random causes, After this period, the hazard rate again 8 increuge und therefore more fallurgs are also observed. This is