GENERAL CONCEPTS |. Introduction to measurement and instrumentation—Measurement-Instrumentation; Significance of measurements; 1.3. Standards of measurement, 1.4. S.|. units and conversion 1.5. Mechanical measurements; 1.6. Methods of measurement: 1.7. Modes of measurement; Generalised measurement system and its functional elements; 1.9. Instruments-Introduction— sification of instruments-Factors relating to selection of instruments-Functions of instruments; 9. ingut-output configurations of measuring instruments and measuring systems; 1.11. Applications Imeasurement systems-Highiights-Objective Type Questions-Theoretical Questions INTRODUCTION TO MEASUREMENT AND INSTRUMENTATION 4. Measurement the act, of the result of a quantitative comparison between a predetermined standard n magnitude) The procedure and apparatus employed for obtaining the comparison, provable; the procedure for this is called calibration. Probably it has rightly been said: “whatever exists, exists in some amount”. The determination of the amount is what measurement is all about. If those things that exist are related to the practice of mechanical engineering, then the determination of their amounts constitutes the subject of mechanical measurements. Measurements provide us with a means of describing various phenomena in quantitative terms. The measurements are not necessarily carried out by mechanical means. e following quantities are typically within the scope of mechanical measurements: Pressure — Temperature Displacement — Fluid flow and related parameters = Acoustics — Mass, length, time etc Tn several situations, however, the above quantities are not measured by purely mechanical means, fnvariably measured by electrical means by ransducing them into an analogous electrical quantity. Jn moder technology, the measurement of mechanical quantities involves the use of electrical and ironic techniques. Measurandl. Ge physical quantity or the characteristic condition which is the object of measurement Batic” or “process variable”)JMENTATION rs AND INSTRUI ecANCAL MEASUREMENTS AND NST INTRODUCTION TO MEASUREMENT AND NS measurand may be: ; Nene h, mass and time; acceleration, Fundamental quantity, €.8 ced, velocity. — Derived quantity, ¢.g.,speed i Standard presse cula Known quantity) rd andis used to designate the particular (Known quantity) The word measurand is use ; that physical parameter being observed and es ns is the input quantity to the measuring z on =. of measurement produces @resul ianee a fundamental measuring proce: rsa . 1.2) gE The i easuring system is known : — Whereas the input to a measuring ‘ as measurand and, the output is called measurement (See Fi Measurand x Fig. 1.2. The basic requirements for getting meaningful result of measurement are: WThe standard employed for comparison purposes must be accurately defined and seh commonly’ acceptable. {The standard must be of the same character as the measurand and usually, Prescribed and defined by a legal or recognized agency or organisation e.g, Invi But Standards (IBS) or the Imernational Organisation of Standards (ISO), \ “_lne sparais used and method adopted for the comparison purposes must be prot with time in a continuous manner over a range of? As _For example, the speed of the automobile, = m as it starts from res a ma uring its motion. A sensor that responds tc fom rest has some magnitude at eve... Se ety Will produce an output signal HAVE py but ot aah ® ‘Several measurands of interest vary Setigare a @ value at every instant of time, becomes a digital Sg@ i a single spec ni scrote numbers each correspondi rhe analn® : ‘ specific instant of time. anding to the value of the ‘ 7 ~ ‘The digital signal contains no ii than sample times no infor 4 : mation about the value of the anatog signals at 8 Such as , for the purpo HOWE by coe ea Stain: tompciatures. nurginow BE Mor a tae et POS Of signal conde ting an analog ao obiained: ME Aotion read out ofr eo equivalent oftenly following distinct ad!4.1.2. Instrumentation he technology of using instruments to measure and control the physical and chemical properties of materials is called “Instrumentation # When the instruments ate used for the measurement and control of industrial manufacturing, conversion or treatment processes, the term “ * ‘process instrumentation) i8 used. ° When the measuring and controlling instruments are combined so that measurements provide impulses for remote automatic action, the result is called a control system SIGNIFICANCE OF MEASUREMENTS The significance of measurements is discussed below: y/ Measurements play a very significant role in every branch of scientific research and engineering processes which include the following: ~ Conirol systems; ~ Process instrumentation; ~ Data reduction, 4 The whole arew of automation oF automatic controls is based on measurements. The very concept of control is based on the comparison of the actual condition and the desired performance. The eaactness of error depends on the precision and accuracy of measurements made, ‘W The measurements confirm the validity of a hypothesis and also add to it its understanding. This eventually leads to new discoveries that require new and sophisticated measuring techniques, wf Through measurements a product can be designed or a process be operated with maximum. efficiency, minimum cost, and with desired degree of reliability and maintainability. HEM stanparps oF MEASUREMENT As the measurement is a process of comparison, therefore, regardless of our measurement method, we ‘must employ a basis of compurison—standardiced units. The standard must be precisely defined, and because different systems of units exist, the method of conyersion from system to system must be ‘mutually agreed upon. ~ Nis of significant importance that a relationship between the standards and the read out scale of each measuring system must be established through a process known as calibration The standards of measurement are often classified as follows: WPrimary standards secondary standards Working standards. «measurement standard 1 the physical embodiment of the wii of measurement ax well as that °F Us -submudtiple valid. “Primary standards: he ghetto of ener bake salt ore Seri wb 5 cated @ primary standard) * Those standards essentially ure copies of intemational prototypes and are Kept throughout the “orld in national standard laboratories and institutions of similar standing.otese Quite sable considering the advan reaching and for sciey pendent Indep defined units given in Invariant No relative but finite the primary standards, the following points are given due con 1 up the primary stands ‘Quantity White setting x Long time stability of the material = y/Low temperature coefficient i ia oy Ae tee its characteristics with changes in humidity, tempers gy | tne cleric cuent ‘hermody namic tem uminoas intensity ipo detericration ‘environmental conditions. wef Rigidity of construction Ww7 Accuracy of machining. ye Most sophisticated techniques to be used fr measuring physical dimensions whiheg the accuracy of the standard. ee amount of substan ‘These standards have the “highest possible accuracy” but are very expensive toon) _ Derived units mn ‘ Z Secondary standards: Quantit u Gre see " = are the reference calibrated standards designed and calibrated fm | ¥% Periodically fo the national standard laboratories for their calibration (Ie | wwe: aime aes depends upon the accuracy and the type of standard being maintained) pe RT, 4 chet tn ita as ny etn and the instal A | Sm + Working standards: | me Lee nr ae activity (rain lites stele . < sae one order lower than that of the secondary dt | specific volume i Carry a the measurements MUSH S18 used by the workers and rechnicians mime |_‘wreoace EEG s.. units ano con VE aaa ity A: Introduction to $1, Units sou SSCTORS | SAL, the International system of UNIS is. di ean livided into three classes: t oe Se point of atitrary, because itn ne oe iypelow: ngths in vacuum of the radi, Md Sd of the Krypton-86 atom” at prototype of the kilos Coulomt \Was Metre. The me } A Somespaning 19 yes ig the mass oF 1PE internation - ; a ns passing ¢ \u Kilogram. OF Bilogram is eat eek periods ofthe eiaton comeipng, Pain * yrSecona TRE second 38 IE duration 1 els of the round state 0 ini 138) ere transition berween the : constant current which, if ipaintained ie be straight pang “Me yy-Ampere, Ove ampere is IH °° gible circular oxe-section and placed one mete ondactors of sft LEME Tose conductors & forse ‘equal to 2 x 10°? newton perme comacutum, would produce Between a of length. oe ae modynamic temperature of the triple hm. Thi weet Kelvin, The Kelvin is the fraction 3.46 of thermody! Bees of water. Le Siemen, 1 in the perpendicular direction, of a surly resistance mho (®). yef Candela, The candela i the uminans intensity ack body ata temperature of freezing platinum under aN Yai, The of potenti Hertz, Th square metre of al * 600,000 of 101, 325 newton per square metre Mle Phe mole is the amount of subs Shaities as there are atoms in 0.012 KB, mtities must be specified and may be atoms, tance of a system which contains as many elemess ‘carbon 13) When the mole is used, the elem [Link] riolecules, ions, electrons, other partis} power) eqi 5 specified groups of such particles. % (ier Radian. ‘The radian is the plane angle between two radii of a circle that cut off on the cf P ‘an are equal in length to the radius. : y-Sieradian. The steradian is the solid angle which having it in the € & g its vertex in the centre of a Tp, t cuts of an area of the surface of the sphere equal t alien ee wees phere equal to that of a square with sides of engh# Some other definitions: es ZN i itil, nearer Pe ton (N) sa derived unio force and is define asthe unit of force Pet eagle logram gives it an acceleration of one metre per second per ¢° Pi acceleration due to gravity equals 9.81 m/s*; i Py Joule, The joule (J) is a deri ee ns 281 news ee the work done when a force eee of energy, work or quantity of heat and is eee eWION Acts SO as to cause a displacement of one me! f] 6 Fo 6, For is defined as the energy in nucl i afte teny cime ym cine cease a potent cont TeV = 1.6021 x 10-19 eiraniss Watt, The watt (W) is a unit of power (ie, rate of doi sees Win P jing work) in Powers seve fer in watts = Wotk (or energy) in joules is time in seconds ITwhi (34 is a non-gravational system of units It clearly distinguishes Berwcen the units ‘of mass and sor ‘weight (fore) which are kilogram and newion respectively. ‘WAI the units of the system can be derived from the base and supplementary unis ned /the decimal relationship between units of same quantity ‘makes possible to express any small 2 ‘'s or large quantity as a power of 10. Stan so te, eee ‘whit «6. Foe any quitntity there 1s one and only, one Sl unit. For Shite Je, Joule isthe unit of ener ae Fecal gy schaleal bon, enesiloal, cleat mod puehes HONEY kWh will also ‘S| UNITS AND CONVERSION FACTORS: Thus 1 watt equals 1 joule/sec 1 kilowatc-hour (kWh) 1000 watt-hours = 3600000 joules Coulomb, The coulomb (C) is the derived unit of charge. It is defined as the quantity of electricity passing 4 given point in a circuit when a current of 1A ts maintained for 1 second. Q=16 where, 1 = current in amperes, and 1 = time in seconds. 1 coulomb represents 6.24 x 10! electrons ‘Obm. The ohm (2) is the unit of clectric resistance apd is defined as| (ie resistance in which a constant Siemen, The siemen is a unit of electric conductance (i.e, reciprocal of resistance). If a circuit has a resistance of 5 ohms, its conductance is 0.2 siemens. A more commonly used name for siemen is mho®) Volt. The volt is a unit of potential difference and electromotive force. It is defined as the difference of potential across a resistance of 1 ohm carrying a current of 1 ampere Hertz. The hertz (Hz) is a unit of frequency 1 Hz = 1 cycle per second. Horse-power. It is a practical unit of mechanical output. BHP (Brish horse power or brake horse ver) equals 746 watts The meric horse power equals 735.5 watts. To avoid confusion between BHP td meine horse power, the mechanical output of machines in S.J. units, i expressed in watts or Kilowatt €.Salient Features of $.1. Units The salient features of S.1. units are ax follows: WA is « coherent system of units, 7. product or quotient of any two base quantities results in 1 anit Scealtent quantity, For example, unit length divided by unit time gives unit velocity. ‘Ziv is a rationalised system of units, applicable to both, magnetism and electricity. continue 10 be used as unit of electrical enersy. Advantages of SL. Units: WAopits for many different quar fing an absolute system, it several expressions in physics Physics and engineering. ties are related through a series of simple and basic relationship nvotds the wae ot factor gen acceleration duc 10 gravity ved engineering ‘which had been a nuisance in all numericals inTATION MECHANICAL MEASUREMENTS AND INSTRUMENTS | > 51 UNITS AND CONVERSION FACTORS ing rationalised system, it ensues al the vantages of rationalised Mx Orca eonaen: ‘nagnetism, electrical engineering and a i Toute is the only sole unit of energy of all forms and watt is the sole unit of power is saved in calculations mi eect system of units and involves only decimal co-efficients. Hence convenient and quick system for calculations. 6 Ineletcty, al the practical unis ike vot ohm, ampere, henry, farad, coulomb, je accepted in industry and laboratories all over the world for well over a century hays a, \ absolute in their own right in the S.1. system without the need for any more practi Disadvantages: 1 Tie non-S1. time units ‘minute? and ‘hour’ will stl continue t0 be used until the clots » watches are all changed to kilo seconds and mega seconds etc. 2eThe base unit kilogram (kg) includes a prefix, which creates an ambiguity in the we multipliers with gram, acteria in milk is di bacteria elimination can be measu red indirectly by ect dePendent upon its temperature. TS? Y measuring the r 4 ~ In indirect measurements an emy aa wirical relatio : J faualy made and the rests that are despa St0blshed between she meas ‘When the primary purpose of making quality should be measured directly, Ho indirect measuremem should be made, Measurement is t Wever, in case di ‘0 determine quality of a product rect measurement is not possible:qpbe pressure g8UEE penn py means of an CICS Thong, J, St ure bY # fe if static Pres® ing shat Jace is called 1 GD) The measurement 0! fee fa ros rand takes ple he tray, . (ii) The measurement of 82 $F atin 0! w (The unit of a measuring sYS!™ include the whok Ge unit of tems and they rag or transaton > © Majority of measurement mechanical, electrical, PC Measurements may also be divi \ Ae Contact type In this ease the sen: k <2-Non-contact type. Here the sensor Non-contact ‘measurements jnclude optical TE! GENERALISED MEASUREMENT SYS ELEMENTS — —=8 Fig. 1A. shows a measurement ysiem (generalized) with different ‘components (called elemers § ; tertiary a7 end electro-pneumatic instrumegy amalic eet quo coregorics: uring deviee convacts the controlled sor of the men ine controlled medium does not contact 1, radioactive am M AND ITS FUNCTION, id others. The various elements are: 1, Primary sensing element. It is an eles clements sense the condition, state or value of the proce: 2a : : of energy from the measurand, and then produce an output which reflects this conditions or value of the measurand. 2, Variable conversion of transducer element. This element converts the signal from one ps ‘form to another without changing the information content of the signal. 3 Stag 3. Manipulation element. This element operates on the signal according to some math = g ‘hout changing the physical nature of the variable. ment that is sensitive to the measured Variable. Thessy sss variable by extracting @ sly we aoe ae © Data transmission cement This element q mation conte 8. Data presentation element. Thy. nis, {from the manipulaion etemenss, "= PPOvides a display record or indication 4 "8 1.5 Stages ofthe general measurement i The a a 4 4 general measuring system COnSIstS Of the E : ‘ansimits the signal from one locatiot ®” |. BournSTATIC AND DYNAN CHARACTERISTICS OF INSTRUMEN, = * ' & intent itatin etinrneroicn. ben N 1 2.2, Delinitions relating te measuring instrurniant®, 4 : nt, aersdcrrecion stato calibation Mange fd span Boa nanan ety mAccuraoy and praoiplon-Bansitivily-Linem sit TH OHI4 op 5 Tene on de tne-Londi etfcke-Noine, 24, Dyninks Charnclorats pag =~ Teenonse-Dynamie characarietce of 0 mmeaburemant syslom/Dynanic anatyie if Mamie Z Syaame-ZorD, frat and second order ayatame=Zaro-order systems Flreb order” eystenns tim Stier systems-First-order ayatam rosponses-Second-order systamn responens, VicrHed Exp Highights-Objective Type Questions Theoralical Quentions-Unaolvad Examples » ' ww ERB intropuction 4 ‘The performance of an instrument is dexeribed by mewdis of quantitative qualities, terme characteristics; these charucteristics having heen broken down inte the sub areas of static charwetere ; Se : and) dynamic hiiructeristios— . The static characteristics pertain 10 a system where the quantities to be measured are coma w vary slowly with time, Performance criteria based on dynamic relations (involving rapidly ® quantities) constitue dynamic characteristics. W { we) ~ ‘The static characteristies, ima real seme, also influence the quality of measurement under conditions, but these characteristics (static) show up as non-linear or statistical effects ia “we linene differential équationa, giving the dynamic characteristics, These effecrs would mat = O* differential equations analytically unmanageable and so the conventional approach ®* oe ei ie nei separately, Thus, even though these effects influence en { wiour, the differential equations of d si friction, backlash, per iia ae” ee ora ores 461 } ATi overall performance of an instrument is judged by a ommend el static and dynamic characteristics. cr cca gee me eal DEFINITIONS RELATING TO MEASURING INSTRUMENTS Some ae terms used in connection with measuring instruments are discussed below: True or actual value. The actual magnitude of a signal input 1 system wt? we be approached and never evaluated is termed as true or actual value, Indicated value. It is the magnitude of a variable indicated by a measuring insrrumen! fate __ditimrT generate i MEO gM serestis more OF LESS COnstan, _ Menta (€8 OCeUraC) er ot Hefescription of MeaSUrement ‘og y We hay Hometea whe gn. (2.4) ma 23.4. © Theenorn 2 e ae means yields the “true valu e eae soon vyrigh a measure oe oe isthe des orrection: ‘accepted engincerin® sandards amount by which the result differs fror ol mv ain amu Gee The difference be rps semi ales We a Geviation duet the vig! average of an finite PUT is sible ro determine experimentally the true value. Then : 4 factor tends 10279) sales ne Toading effects. lags and mechanical ees ted any measurement SEM ue the foading efers: 148 chanical problem, error. but opposi ' Frsteresi, nse pic €16) “Accuracy ' pecuacy ofthe measured signal depends WPON the following factors . (@ Point «Variation ofthe signal being measured (ii) Percent x} ‘nisi aocuracy of the instrument itself (ii) Percen *f 'e Accuracy of the observer; (i) Point a ‘Wher oot the quant is Being uly impressed upon the instrument Examples Errors: (a) Agive ‘The accuracy of any measuremet b) As 7 i (6) Ascal k the trve value of a quantity, as ismeasured in terms of its errors. Since it simp Se cena es ili ep imation ofthe “rue value” obtained NY 8 Cer eon aca pea caer and conditions to which corresto® r and the tre value ofthe quantiy i ralue” of the quantity. The difference betweet™ N The absolute ty is known as 4 value of error does not indi static error. Eqn. ( S not indicate preci: 1 crs precisely the accuracy of measuremen perso fi (iii) Perce value ¥ V,=T alue of quantity, ; = True value of quanti ie quantity, ~ He Gene + = Slat (or abso = Rel lute) error i ir icone of measurement, and‘Then, and, ‘Absotoie etree Tie value ‘or, Sage relative error, % E, = E, x 100 ah Me {From eqn. (2.1)] {From eqn. (2.2)} v, yee 1 14, AZ) However, when £, is small, E, << 1, Eqn. (2.4) may be written as: V,=V_(-E,) (25) «¢-‘Theerror may be cither +ve or-ve. For +ve static errors the instrument reads high whereas it reads low for—ve static errors. Correction: The difference berween the true value and the measured value of a quantity i called static correction (C,) ies 1 Vp (== E 2.6) Eqn. (2.6) indicates that the static correction of the instrument reading is ofthe same magnitude as the extor, but opposite in sign (ie., C,=—E,) “Accuracy” may be expressed as: (@ Point accuracy, or (ii) Percent of scale range, oF (iii) Percent of true value. (@ Point accuracy. In this ease the accuracy of an instruments stated forone or more pointsina range. Examples: (a) A given thermometer may be stated fo read # 0.2°C between 50°C and 100°C: (6) A scale of length may be read within + 0.2 mm. (ii) Percent of scale range. Here, the eror is caleulated on the basis of maximum value of the scale. % Evo lea ae Jes (21) qn, (2.7) refers to the intrinsic accuracy of the instrument itself and does not include procedural or personal performance. Giiy Percent of true value, In this case the absolute error of measurement expressed as Sage of true value of the unknown quantity: hen % Ervor= feats 100 (2.8) — Here, with a drop in true value, the size or error decreases,jpility oducibi nd ReP* nd reproducibility sity 2 rability repeat padings when the sam xs of omput readin = ene’ emt conditions, same insiryt' ase measure ms out. we with # x vaimained throughe ¢ 2.3.7. Accuracy ; Accuracy. The it readings for the Same inpy yy | measured s called accun utp! closeness of nstrument, location, condiyg My, "= val the er, measuring I” nition, dead» pret: 4 men Precision. 7 sinument OUPUL OVEF 0 Period of ting nnure of HE Follewha ecus -ondiions OF load. pein © bee luces same readings at different ting ampere, but instrument sssroven rife iit repro! F enbetican i in messored variable the following factors: bape | Toe iftrmay be cased BY ‘in some parts of instruments and systems (ity Take the eass (yi etn sree decor me \ (i Wearandiear, Soom (a) Mechanical ee accurate, : | Tey SE ea mgt ils ese : (6) Thermal emf : bas eemples ahah 3.8. Sensitivit : © Drift occurs in “resistance thermometers and thermocouples” due to the contamnaehic ratio of the ‘and a change in its atomic or metallurgical structure. b \ | * Daitimay oceurin “low meters” due to wear and erasion of the orifice plate, = : “ere When thee | Dri maybe clasifedas follows: 1 Zero di i reocindit: [Link]; 3. 2onat ai, Dri called cero di a us Static sex oe ae °F instrument calibration gradually shifis over bY i 2 Span arin: SMS#10F lippage and can be corrected by shifting ark j he calibration ial | oy "ward chan 5; a ie oat gnh hg tng oe itis called “span (ors © Seastivity h it cera Portion alk B is am und at instrument, itis called “° ft "Hts ey oy ee © "ble quantity since it is rare * aN be nvesti‘STATIC CHARACTERISTICS. 4 2.3.7. Accuracy and Precision ‘Accuracy. The closeness with which an instrument reading approaches the true value of the quantity being measured is called accuracy. Accuracy is determined as the maximum amount by which the result differs from the trie value, Accuracy of an instrument is influenced by factors like static error, dynamic error, reproducibility, dead 2008 Precision. The term ‘precise® means clearly or sharply defined. Precision is a measure of reproducibility of ‘measurements, It is usually expressed in terms of the deviation in measurement. ~ Significant figure is an indication of the precision of measurement. Significant figures convey actual information regarding the magnitude of precision. Following examples will illustrate the difference in meaning of the two terms accuracy and precision: (9 Letus consider the case of an ammeter having high precision which can take readings to 1/150 of an mpere, but has a wrong zero adjustment. Now we can take readings to 1/150 of an ampere which are consistent and clearly defined since the ammeter is precise as ever. However, the readings taken with his ammeter are not accurate, since they do not confirm to truth because of faulty adjustment of 2810. (ii) Take the case of a micrometer which is normal in every respect but its anvil is displaced from its rue position. When the readings are taken by this instrument, these would be clearly defined and consistent ‘and we would say that micrometer is as precise as ever. However, the readings do not conform to the truth as anvil is displaced from its true position. Thus, the readings taken by this micrometer are nor ‘accurate. ‘© Iris worth noting that precision is a necessary prerequisite 10 accuracy but it does not guarantee accuracy 2.3.8. Sensitivity The ratio of the magnitude of output signal to the input signal or response of measuring system t0 the ‘quantity being measured is called sensitivity. Itis represented by the slope of the calibration curve if the ordinates are expressed in actual units. ~ When the calibration curve is linear (Fig. 2.2a), the sensitivity is constant. However, if the calibration curveis non-linear (Fig. 2.2h) the sensitivity is different at different points, being the slope of curve at ‘Yarious points Static sensitivity, in general, is defined as: Infinitesimal change in output _ AA, Suucsensitvily (9) = 7 nitestmalchange in input, AA; Sal aA, © The sensitivity of an instrument should be high and thus the range should nor be high in comparison 10 the value being measured. ® Sensitivity has a wide range of units, and these depend upon the instrument or measurement, system being investigated. 2 Similarly, Inverse sensitivity or deflection factor or scale factoSTATIC AND DYNAMIC CHARACTERISTICS OF INSTRUMENTS craracterionics) * This type of noise may be eaused by the power supply (to the amplifier) which may have, ripples or random deviation: © 50 Hy power supply and the uarmoni¢s contained in it are one of the common sourees for conducted © Ir can be reduced by using filters in the leads to trap out the noise (iii) Radiated noise Radiated noise consists of unwanted signals (relating to electric or magnetic fields or disturbances in the environments around the amplifier) which are radiated into the interior of the amplifier © The electromagnetic impulses radiated from ignition wiring of spark plugs are one of the common sources of radiated noise. © Iyeain be reduced by proper shielding DYNAMIC CHARACTERISTICS When an instrument 1s requited to measure a time-varying process variable, one has to be concerned with ‘dynamic characteristics”, these characteristics quantify the dynamic relation bepween the input and output, 2.4.1. Dynamic Response The evaluation of the ability of a system to faithfully transmit and present all the pertinent information included in the input signal and to exclude all else, is called the response, The behaviour ofthe system when inputs vary with time (i.e, inputs are dynamic in nature) and so does the ouput, is called dynamic response. Such a response in measurement systems is found to occur in industrial, aerospace and biological applications “The dynamic inputs are of two types ( Steady state periodie; (ii) Transient. The “steady state periodic” quantity is one whose magnitude has a definite repeating time, whereas the time variation of a “transient” magnitude does not repeat © The response of a measurement system subjected to a time varying input can be divided into the following two pans: (i Steady state response; (ii) Transient response. “Steady state response” is simply the response when time reaches infinity “Transient response” in measurement systems, is defined as the part of response which goes to zeroas time becomes large. © Periodically varying inputs may be: A purely sinusoidal input like the A.C. voltage with a frequency of 50 Hz or ~ A-complex time varying repetitive input which is @ combination of the fundamental and harmonics. ‘When the systems are subjected to periodically varying inputs, they exhibit in theiresponse a magnitude 4nd phase relationship whieh is different from that of the input signal because ofthe energy storage elementscontents of the input are not ry jarmonie distorted b pe input THe bare. The output is distorted bod ; semasion of tei ne np snot aft ean phase relations p ‘The output is ide and il inthe utp bot phase relation sD of magnitude as vel surement System urement system are amie characteristics of # meas ‘The dyne es Desirable 1. Time dor 1, Speed of response | Undesirable In this thei Desirable Function of t “Undesirable isanalysed sponse” or “responsiveness” is defined asthe rapidity wihyy Signals van es . 4 jonse. The “speed ofr aes set ire em responds 1 changes in the measured quant following si — aie of a measurement system 10 che “Measuring lag. Itrefe oe % OSe carey Bm: (i) Ran ag is caused by conditions such as capacitance, inertia, or resistance. ee “The measuring lags ae of the following two types: ia (@ Retardation type lag. In this type of measuring lag the response of the measurements: > Frequen ~Pegins immediately after a change in measured quantity has occurred. Gi) Time delay type lag. in this case the response of the measurement system begins after ai ‘ime after the application ofthe input. the measured quantity varies at a fast rate, the desl hhasa severe adverse effect on the performance of the system. The measurement lags of thst This type of the system independen cl © Fr are very small and ate ofthe order of a fraction of a second and hence can be ignored. pits GS Fidelity. ris defined asthe degree to which a measuremen tem indicates changes in the mea not hou any mic error, [refers to the [Link] the sy ss _Sei ete dence Me [Link] the oulpil! Standard te A Dynamic error. The dynamic error, also. true value of the quantity changing w with time y it static errors assumed (see Fig.2.5,) “nde vale indicated byt ‘The most co Called “measurement error”, isthe difference betwe 1¢ measurement syste! 1. Step funei Refer to Fig Itis ma of a step input of 80 bar to a and finally comes to rest at 81.3 bar. Determines,” B86. the pointer swings to pressure of 5! Where [Link] () The ' nae oe es ofthe gauge reading and express as @ Be signa i percentage error ofthe gauge Semtage of the final reading. = Ramp or Solution. i) Ove BENIN (9 overshoot ofthe yauge reat In this case, Percentage overshoax « 12 yyy a 813 Wor = 813= "ercentage error = = 80 Where, yisERRORS IN MEASUREMENTS 44, Introduction; 3.2. Limiting errors; 3.3. Types of errors—Gross errors~Systematic errors~Random errors; 3.4. Sources of errors; 3.5. Statistical analysis of test data~Slatistical averages-Dispersions from mean-Best value from a sample of readings-Normal or Gaussian curve of errors-Probability tables-Odds-Specification of measurement data—Confidence interval and confidence level-Probabllity graph paper-Rejection of test data-Chauvenet's criterion-Method of least squares; Highlights~ Objective Type Questions-Theoretical Questions-Unsolved Examples, ma INTRODUCTION Errors and uncertainties are inherent in the process of making any measurement and in the instrument with ‘which the measurements are made. The study of erroris important as a step in finding ways of reducing them, and also as means of estimating the reliability of final results, The accuracy and precision of a measuring instrument depends upon its design proper selection of material for each and eyery part, and skill of manufacturing. fa LIMITING (OR GUARANTEE) ERRORS {In most instruments the accuracy is generated to be within a certain percentage of full scale reading. The ‘manufacturer has to specify the deviations from the nominal value ofa particular quantity. The limits of these deviations from the specified value are defined as Limiting errorsor Guarantee errors The relative error is defined as the ratio of the error tothe Tf, specified (nominal) magnitude of a quantity: Q, = The specified quantity or true quantity of a component, and Q, = The actual or erroneous quantity; Then, enror = 0,- 0,(=30) and, Relative limiting error (also called accuracy), g = 2-2 _ 80 5° 50) Percentage limiting error, A _ %E, =E,x100 4 component has a relative limitin i n= OLE ne is etter +€E,, then limiting values of the component, ae ;,). Thus Fae a menitude ofa resistors 10002 with limiting enor of 4 ohms, then he i 96 ohms and 104 ohms. In other words theor of = 1-5 percent — nr flow being measured i ‘\ e=o 1 Hence, Percentage lomiing error WE sow fu cale det ito percentage of ull scale deflection Ne me sco 5x 10 mils the flow meter is porate 1s me Afecale valve of 125» 10 why net 15 \x500 =27.5W MJoubled. This indicates that the percent 100 bcrefore, while selecting insrument, ing is 50 W may be $0 2 75j, ¢csured should not leo the lower thi oe is Ric scale should be used s75W(Ans) sultant error of a device composed e EB xio0 =15% sie Relaive err, rot. The limiting ero of th . ee age of true value, then, {the vanous components are com ow suppose the eros specified as percentse 15 5 W. When two ot inore quantities, each h Magainde of error = # 55,50. = 20.75 W. and timing eno of he ebination Th 100 vencticin of the final result i in the forme “The meter may read between $00.75 or 49.25 and 50.75 W. (Ans.) i) When the final results eithet Ema Leh ged dyn dy = 6 The guaranteed accuracy of a flow meter working on thermal principles is * 3 perc’ Qa of 25 x 10 wl. Ifthe flow measured by this meter is 1.25 x 10° mls, casi ®* hen, (sum oraifference)= perce. Comment ypon the results. 82. BERIEN Given - Guaranteed accuracy of flow meter, ~ 24) E, = £3% (of full scale reading) " iY Equation (3.1) shows that the Resets 2, =25% 10m Ic yuki che nti ss Ping ‘ 10° mis, ‘moter = "3 9} Gi) When the result is either pi Mimitng ero Percent): ee In this ease ‘The magnitude of lien OF limiting error at full Scale ofthe meter is or =e x0, or. = £00325 x 10-4 Then, ‘The magnitude oe Now being nS 10 ms © The relative error. Uh fo ‘Measured ig 1.25 y¢ 10 mis: nother words the relarive limERRORS IN MEASUREMENTS: LIMITING (OR GUARANTEE) ERRORS The value of flow being measured is, = O(1 + F,)=1.25« 10%) £0.06) = 125 x 10% + 0.075 x 10% nis Hence, Percentage limiting error, 0.07510 %E = 25x10 100 =26 % (Ans) Comments. From above results we observe that when the meter reading is full scale value of 25x 10-°m's the flow meter is guaranteed to have a limiting error of + 3 %, but when the meter is reading its half scale value of 1.25 x 10“ m'/s, the limiting error is + 6 %. Thus when reading is halved, the limiting error is doubled. This indicates that the percent limiting error will be greater if smaller values areread by the meter. Therefore, while selecting instruments, special care should be taken as regards the range; the values to be measured should not lie in the lower third of the range. Hence, as far as possible, meter which will read up their scale should be used. Resultant error of a device composed of several components: ‘A measuring instrument/device/system may be formed by combining several componen's, each having its fimiting error. The limiting error of the combination can be computed in the following ways depending on how the various components are connected in the total system: ‘When two or more quantities, each having a limiting eror, are combined, itis advantageous tobe able o compute the limiting error of the combination. The limiting err can be easily found by considering the relative increment of the function ifthe final result is inthe form of an algebraic equation. (@ When the final result is either sum or difference of outputs of various components Let, 94.42.95: 9, = Outputs of each device, and Q = The final result; ‘Then, Q (sum or difference)=4, 4,4 4,+ and, 8 _4/% She ds B1) Q Q4 2% Equation (3.1) shows that the resulfant limiting error in this case is equal 10 the sum of the products Jormed by multiplying the individual relative errors by the ratio of each term of the function, (Gi When the result is either product or quotient of output of various components: this case, 0 = 4,%4)% 45% % = 9+ 4,4 43% Iq = 4,%42* 43% 44* 80 (i,m ,On, Om 32) Q hm & Gq nother words the relative limiting error of system is the sum of relative errors ofall devicevcomponents.~ aap om ay Whe the fal is anh #0 srisenses hh, f 2. [Limiting extor in the m ‘Then _— sve ine flowing rnin 265 22 4%; Ry = 45 D2 4G poe ae she are enter Br % arts adhe ere Determine the, following ¢ ; ie resent resistance a 4) Limiting vale of Vl 4 A uy Perce iting e707 series combination of resistances R= ETB ven’ R= 254M $2+4% Jor the unknown resistance). 4, =45 044% iting ale of the resultant resistance: ' SEM Given &, “The unknown resistance is gh ‘The values of the resistances are = where R Determine the magnitude of th 0205 4 R, = 254— =25 1 pyres 352100 Ry: ‘The relative limiting error of 4 = 654—-x65 = Thad 652.62 . Ry R= 4544 = 454-7 x45 = 4521.80 L % ‘The limiting value of resultant resistance, es . Limiting error in ohms re al Th i oa ie, The guaranteed valu on “t ee e254 DADs) 4004 66) 9 (Ans) Tr ofthe combination nbination of resistances: Zn cantilever beam of length. Wi sing an LVDT of accuracy = 0. nad their values are 45 ‘mm, 4. -2 Se. Ifthe deflection noted Given force. where, E = Modulus 1 = length of SEMEN the limiting errP ‘We know that, Rea ‘s. Limiting error in the measurement of resistance, ae (SP 428 R Ti = =(1.242x08)=22.8% (Ans.) The unknown resistance in a Wheatstone bridge is measured by using three known resistances ssuch thats RyRy x where, Ry = 50+ 05% O; R, = 500 + 05% 2; Ry = 40 05% 0 Determine the magnitude ofthe unknown resistance (R,) and the limiting error in percent and in olims (for the unknown resistance), Ry Given: R,=50£ 05%; Ry =500+0:5% 2; Ry=440205%2 ‘The unknown resistance is given by: 500% 440 50 The relative limiting error of unknown resistance is, (Oe oe RR = #(0.5 40.5+05)=21.5% (Ans) 4400 Q (Ans.) R Limiting error in ohms = ix 00 = +660 (Ans.) Le. The guaranteed values of resistance are between 4334 (i.e., 4400 - 66) © and i ee ) 2 and 4466 (i.e. A cantilever beam of lens i is ce gt |, width b and depth d is used 10 measure the force F. The deflection i laingan VDT of accuracy £0. mm. The length, width and depth are measured to an ee Geter values are 43 mm, 45 mm and 0.9 mum respectively. The modulus of rigidity is known ‘accurate If the deflection noted was 1.8 mm, determine the error in force measured. be 1 = length of the beam; i REMI the lining erorinfoceFi,a ee ee fellewcine ream: pa TEESSIDE ay where. F = Force ax the end of torgue arm (zh L = Length of sorgue arm ram, B = Number of revolutions daring time and 4 = Time fortes: rans) 10 reveducions: The errors are limiting (absolsae) errors. Shean Dezermine: (@) The magnitude of power. (Gi) The magnitude of the limiting error in the computed power, BET Giren: F£= 426 2002 bg: L= 382212 R=11922 4) The magnitude of power P : 2RX9RIXFIR = ~IGiver) ‘i Limiting errorsin the computed power: Total per unit error in computed power.Limiting error in computed power SP = £0.017 xP = 20.017 x 1.993 =0.0330 kW The result is written as: P = 1.993-+0.0339KW (Ane) Emo Bi (a) How are “known errors” treated ? (6) During a text the current wax measured as 26.5 A flowing in a resistor of 0.12 2 Later on it was discovered that the ammeter reading was low by 1.1% and the marked resistance was high by 0.25% Determine the true power as a percentage of the power that was originally calculated. (a) Known errors. When the error of quantity or an instrument is known, the effect ofthis enor when combined with errors can be computed in a manner similar to the combinations of the limiting errors. However, the difference is that in case of known errors the signs of relative errors are given and must be preserved in the calculations, (8) True valve of current !=265(1 +0.011)=26.791 A True value of resistance R = 0.12 (1 ~ 0,0025)= 0.1197 9 ‘True value of power = FR = (26.791) x 0.1197 =85.916 W ‘Measured value of power = (26.5) x 0.12 = 84.27 W 85.916 ‘True power = %100 = 101.9% (A — True power 199 = 9 % (Ans) Measured power 27 Power,P = FR = —1__ =1019% (Ans) 100195 Cm (Observational errors 5 i Siu Yeen her ctecbXt Random fie ct ek3.3.1, Gross Errors ‘These errors occur due to human mistakes in reading instruments and recording and cae ‘measurement. Although itis probably impossible to eliminate the gross emors completely. el ome eg, to anticipate and correet them, ‘The mathematical analysis of gross errors is impossible since these may oecur in different sai While some gross errors may be easily detected, others may go unnoticed \Tiese errors can be avoided by adopting 1wo means : _ D-Ammense care should be taken while taking the reading and recording the data ‘Two, three, or even more readings should be taken for the quantity being measured. © Designers of instruments to some extent can lake cure to avoid gross e1 upon the interest and competence of observer s but much would depres 3.3.2. Systematic Errors ‘The systematic errors are repeated consistently with the repetition of the experiment and are cauted by su Bfectas ~ sensitivity shift; ~ 2010 off ~ known non-linearity. © These errors cannot be determined by direct and repetitive observations of the measurand made exch time with same technique. These errors can be located only by having repeated measurements under different conditions or wth different equipment and where possible by an entirely different method. ‘These errors may be insirumental, environmental or observational errors, (2 fistrumental errors: ‘These errors arise due to; — Inherent shortcomings in — Misuse of the instruments, — Loading effects of instruments, Instrument errors are inherent in instruments due to coy struction, calibration or operation ofthe instrumes These errors may cause the instrument to read too low esate oF te or t00 high. Some“examples” of such errors ate : ) Mechanical friction and wear, bac ieldi s rmcnaia: eos ‘Kash, yielding of supports, pen or Pointer drag and hysteresis (4) Constructional faults resulting from finite width of ee pats Tt mon eo nance arn HEE. cenit a te (iii) Improper selection and poor maintenance of the instrument. (iv) Unavoidable physical phenomenon due to fric Assembly errors. These errors may be due to the follo (@ Displaced scale Le., ‘movement. + capillary attraction and imperfect sarefscti™ wing: income iting f the scale zero with respect to the Actual zero position of ae nePointer, ‘The assembly errors do not change wid © with 1 ents. A good A good instrument u: vere and yevnifi 4" unintelligent way may giv ey Of th mment errors : wm. ‘environmental erfors are duc to conditions extemal to th al 10 the measuring fare, humidity, dust, vibrations, external mapneti device ( a external magnetic or ele eg. effects of presuue tefesome than assembly errors, Henetic oF electrostatic fields), These en rayon el hese errors can be eliminated or reduced by taking the following m ures {9 Using the instrument in controlled conditions of ‘originally assembled and calibrated. Pressure, Lcmpsrature, humidity in which it wus (ii) Measure deviations in the local conditions from the calibrat eran he calibrated ones and then apply a» suitable {ii Automatic compensation for departures from the calibrated conditions by using sophisticated devices i) Make a complete new calibration under the local conditions (0) In ease it is suspected that external magnetic or electrostatic fields can effect the readings of the instruments, magnetic or electrostatic shields may be provided, (si) Apply computed corrections. Normally, efforts are made to avoid the use of application of compuied corrections, but where these corrections are needed and are necessary, they ate incorporated forthe computation of the results. XGbrervatonal errors: ‘errors occur due to carelessness of operators. Even when an instrument has been properly selected, ‘ielllyfnstalled and faithfully calibrated, shortcoming in the measurement ‘Becurs due to certain failings on Vo ‘ma & he observational érrors may be due to the following reason lax—Since the parallax errors arise on account of pointer and the scale not rat in the same 4 : vita he amie pines ETRE, we cad clieinate this error by having the pointer and the vale in the same pl j scale reading, and wrong recording of dat ai .¢ reading, lack of ability jnterpolate properly between pa Scnversion of units in bepween consecave anes SM mo-simoltancows obser wersion Bis lnerdependent quantities. “a PP Tenierey to read high or low, anticipate @ signal and se ane estas * Modem cicerical instruments have disital display of uP Ty ome ioe ‘of human observational or sensing powerpp SOURCES OF ERRORS NEGHANICAL MEASUREMENTS ANE INSTAUMENTATION + vrunnt be. predicted 7H) ‘The magnitude any direction oF there eHP ie lay the ayatein; however these exrirn wre ansumiest 1 fil The most common canes np these errr WE {iy Friction i instnument movement (ii) Backtaxt in the movement (iid) Parallas errors between pri 1 (iv) Finite dimensions of the point (0) Hysiores inelastic members (vi) Mechaniva vibrations SOURCES OF ERRORS Hosides the errors so far enumerated (of the phiysienl and a the faulty calibration of a measuring systems the ior anul scale iit aoate ivinions: Faniowl paiwre) and she errors wie p one urine Krom the following surees ay sill jul Information. {i Noise, tris defined as any. slainal that does non convey 41 10 a minimum level through filtering, wavetial sebeenon a anuting, 9/16 Noise aiznalyerrons can be 1, hl (Hi) Response tine, leis defined as the sine taken by the instrumant to shee O3,2% change Inn rete, 10.4 step (apul. This {uctor contributes to tainty of the measurement (iid) Design timitations. tn the design of the instrument, certain inevitable factors euch as frhehww me resolving power lead to uncertainty of measurement Alo {iv Energy exchanged hy interaction, As earlier pointed out, wh the energy requined for uperaing the measuring system is extracted from the measurand, the value of the Latter is altered ko a reser Aeaser extent, This alferation is dependent upon the capacity of the system (») Transmission, During the transmission of information from the prisnary sone indicator, the signal may be attenuated! due to any of the following reasons — TWinay suffer loss through leakage; ° — ICmay be absorbed or otherwise consumed in the communication channel; — may be distorted by resonance, attenuation or dela ‘on various signal component, Y Phenomenon whose actions ase wie” (0) Deterioration of measuring system, ‘measuring system and result in ag ~ Chango in resistance of a circuit element through strain relief — Alteration of thermocouple characteristics th : — Weakening of a permanent magnet, (vii) Ambient influences on measuring systerns (vii) Errors oF observation ind interpretation, The examples of some types of deterioration which ooow fource of ettor jn measurement are + ough oxidizing or reducing. atmosphere— The object of the statistical methods based on laws of chance which operate only on random errors and nol on systematic errors, is to achieve consistency (precision) of Value and not their accuracy. The experimental data i obtained in the following two forms of tests : (i Single-sample test; (i) Multi-sampte test (i) Single-sample test. 4 single measurement (or succession of measurements) done under identical conditions excepting for time is known as “single-sample test”. (ii) Multi-sample test. Ina mal under ample test, repeated measurements of a given quantity are taken different test conditions such as using different instruments, different ways of measurement and by employing different observers. This type of test does not ma measurements with the same equipment, procedure, tech ¢ For getting the exact value of the quantity under measurement, tests should he conducted using as many different procedures, techniques as experiments as practicable. It should be understood that the statistical means which help us to arrive at correct results are only valid for multi-sample tests, 3.5.1. Statistical Averages When several multi-sample observations are taken experimentally there is a scatter of the data about ‘some central value, one of the methods. presenting the result in the form of a histogram (also called a frequency distribution curve), Thus, histogram (Fig. 3.1) is a pictorial representation of the test data ‘where the ordinates indicate the number of observed readings of a particular value, With more and more dats taken at smaller and smaller increments, the histogram finally changes into a smooth curve. The smooth curve is symmetrical with respect to the ‘central value. Gy Arithmetic mean (AM), The most / probable value of measured variable Z = [Link] observed readings | —> (variate) is the arithmetic mean of the _Quantiy —> number of readings taken. We arrive at il the best approximation if the number of Fig. 3.1. Histogram. readings of the same quantity are very large. The mean is computed by summing all the values and dividing by the number of measurements. If gy, 43, 43. 4, be the set of measured values of a quantity, then the arithmetic mean will be given by, ABA) (i) Geometric mean (G.M.). Itis defined as the nth root of the product of n terms. Geometric mean (G-M.)= 919-43 0 4, B.S) (iii) Median. The middle value of a set of an “odd” number of readings, if variables are arranged in numerical order is called the median. For an‘“even'” number of readings, the median is the arithmetic average of the two central readings.WECHANICAL MEAMUMEMEH TR ANID INBTHUMEN TATION Pe RIATINTIOAL ANALYIN GE TRY DATA Wd Maden the vali of dhe warlite ar a fenquenly HH a syrinwrivl dlatribuntion curve; the ree nuenares ef eerlre ell ine pa rane are equal 3.8.2. Dispersion from Moan ey wile demote the eaten ta setae the value are dispersed abit (th they isperston ov stored A Lange disperse eatin th rvenaranie tl gues tw conttol ain! thererore iibeconnes difficult jo estimate te mengtired qanntity with exntideniee un The sioplest posnibte measure af dispersion ty the Hanwe whic be the difference betwren groan tease volien af dan, 16 'Devtutlon, The departure ay the observed reallny from the artthmenlé mean of the group of rei Ws termed as deviation, Lot tye deviation wt readin 4, be 1, and hist of reading 1. be d, e6 Then, ened dant 4 4 an G = Mn) fe sum of deviations.» dy dy 4h, 4d, GDN, “DH DA lg, (a 1a, dy Honk gd nih 00 Ca F044 ack gy) <8 il Thuy, the algebrate: sum of deviations is serv, The above rmathemat result implies that the deviations eun be positive as well as neyative (i) Average deviation, It is defined as the sum of the absolute vulues of deviations divided b number of readings Average dev be expressed as gh = MAU Dd contd | Zdl yy, yu a J The average deviation is taken a measure of accuracy of mea © The average dev measurements, Highly precise instruments yield a low: average deviation henecen rraline on is an indication of the precision of the instruments used in makiog ® (iid Standard deviation. “Standard deviation” (or "mean square deviation) of an intinite m data ix defined as the square root of the sam of the individual deviations squared divided b 4 number of readings Dim“Thus, standard deviation, on (Radeds wedi yaae 139) finite, When the number of observations 1s on pctontoweve te munber of bean i bol @ while if it is less than 20 the greater than 20, standard deviation is denoted by the sym SSmbol used is § The standard devitation of wfinite number of data is given by _ (aed) fe phen Un E210) vel ae N=) Variance. The square of the siandard deviarion'Is-called variance, Variance, But, when the number of observation is less than 20, ( Bes a 3.5.3. Best Value from a Sample of Readings Let, 9,=The best value (approximately the true value) from a set of readings/data, ‘Then, the deviation or individual component error, 4, where, i= This deviation may be + ve, zero or — ve. (i) Sum of deviations ; Da = Da@-a) mim In case of a best value, the sum of deviations is equal to zero. La-%) ia le = 2u4i Which is the arithmetic mean. is ‘Thus the arithmetic it = +e or~ve likely to fey al may be interpreted as the best valve around which the chances of error being tiSie (iy Average deviation the squire Team ine be ies : AT AI $ qh -2m Lait" a gy |i=t ied jation gives: 2 sant the diferent Asall the g? sare cons! s ty i a ‘ Average deviati ial ie %= 5 D4 «. ! sich is again the arithmetic mean. Obviously, the arithmetic mean is the most probable stich is ; ‘value which one can obtain from a set of scattered measurement data. ee { Fis fi may be noted | : - ; #6 Standard d Bs eicencte rev te follwing readings were taken ofa certain physical lth ee 134,138, 1.56, 1.47, 1.42, 1.4, 1.53, 1.48, 1.40, 1.59 mm ‘suming that ony random errors are present, calculate the following: sees) (1) Arithmetic mean, (i) Average deviation, (iit) Standard deviation, and (i) Variance ERM Given: Rea, Far “ading ofthe micrometer Screw Aritimete mean (iv) Variance } i = 1.34, 1.38, 1.56, 1.47, 1.42, 1441 We 000 tha, srithmetig mean, SEnEeEso) aaah oS j Bight di Bighe different su = tite 412, 428, 42: a 4 MOOS +46 +47 + ay + adi Calculate th = M+ 13g 10 | (i) Arithme 41.5641 1H (ii) Average = 1461 ST 149 + 144 4 1.534148 Te (Ans) 10‘ennent H BTATISTIGAL AHALTSIS OF Test OntR ay Avenger Frome eq, (1. the deviations ate aed a iM = tol == 0121 ye hah (Aol = 0001 = 1.80 1Aot = 600000 eA aol 4 0.000 = 1A 1461 0041 We 1Ad = 1a) = 021 J 1.50 140 4 0.060 Je lAde LAbl ey 0019 qd = AAO = LAGI = 0.061 0-H 159 1.461 0.129 Average deviation, «Ul oy glu al mon n 0.121 40,081 40.099 + 0.000 + 0.041 + 0.021 + 0,069 + 0.019 + 0,061 + 0.129 -. 10 = 0.068 (Ans) ‘of ayerage deviations, the signy of deviations are not considered. Ju Standard deviation : Since the number of readings is 10, which is less than 20, therefore equation (3°10) is used for caeulation of standard deviation, a? Standard deviations, s = q/=—y Eqn. (3.10)) “oan! voodoo nO? (COON COUN? + (40060) + OID" + OOM +1297 W- To) 5 = 0.0807 mm (Ans.) Variance, V = 3? = (0,0807)? = 0.00651 mm? (Ans.) MOOR ee Eigh diferent st ie tudents turned in the circuit for resonance and the values of resonant frequency in kHz were (0) Variance : 412 ARS, 423, 41S, 426,411, 423, 416 Galewtate the following : (i) Arithmetic mean, (ii) Average deviation,STATISTICAL ANALYSIS OF TEST OATA <4 i Average deviation : From eqn, (3.6) the deviations are : % 134-1461 =- 0.121 q2~] = 1.38- 1.461 =- 0.081 43 —7 =1,56- 1.461 = + 0.099 da ~@ = 1.47 - 1.461 = + 0,009 qs~ 7 =1.42-1.461 =~ 0.041 7 0.021 4) ~9 =1,53 - 1.461 =+ 0.069 J = 148-1461 =+ 0.019 40-1461 = 0.061 159 1.461 = 0.129 a =. Average deviation, = Eldll,, $linal a0 BeBe") 041214 0.081 + 0,099 + 0.009 + 0.041 + 0,021 4 0.069 + 0.019 + 0.061 + 0.129 - 10 = 0.065 (Ans.) ‘nidy be noted that for calculation of average deviations, the signs of deviations are not considered. wndard deviation : Since the number of readings is 10, which is less than 20, therefore equation used for calculation of standard deviation. Sat Standard deviations, 5 = 4} «o{Eqn. (3.10)) ee & 12D? FOB ODI 4 OOO 4 (DOAN) + 0.021)" + (40.0699 4 (0019)? +0061)" + (0.129)* 0-1) = ease? = 0.0807 mm (Ans.) Variance, V = $= (0.0807) = 0.00651 mm? (Ans.) (i) Variance : Sra Eighe ai i it Mees students turned in the circuit for resonance and the values of resonant frequency in kHz were 412, 428, 423, 415, 426, 411,423, 416 Calculate the following = (i) Arithmetic mean, (0) Average deviation,| i 3.54. Notmal or Gaussian Curve of Errors 2 ee a Pe ERRORS M MEASUREMENTS 2 STATISTICAL ANALYSIS OF TEST OATA Tnexperimental work the random effects are normally studied with reference to the Gaussian cr normal enor genbution Here, followin sumptions are made 4 All ob large number of small random disturbing effects and these small random terrors may be both +e and -ve Thus, there isa strong central tendency. ie, there is @ greater pe {There is equal probability of 4ve and -ve disturbing effects. As such the probability curve of erors inst the magnitude of error will be symmetrical about the zero value ability of small errors and large errors of both +e and ~ ve values bility sates that the normal. oecurrence of deviations from average value infinite number of measurements or observations con be expressed as pete ea of Number of readings at any deviation “d". ce, probability of occur shore P ‘Standard deviation, and = Magnitude of deviation from mean —— ooirOC—~— Fig. 32. Gaussian dstrbutton curve ox The curve is am. The two values 1 value of 6 the hich Phas been plotted as 2 functios g form ofa hist ven figure. Larg sg. 32. shows a Gaussian distribution curve in sbeenp me cal shoot the mean value and may be regarded 5 He Uniti st daha Seyi aya Cylch cares er fazer the curve and hence larger the expected ¢170" 0 If we integrate P between the Limits « valoes of x Because the curve is related (0 ¢. we shall obtain the arca under the curve erween these 9 he probability tha the messurand Les within sme Fangs, RESYS B19) Le. area under the probability curve is unityoa76aNR 08453 ‘Standard deviation for the normal curve: We know that, standard deviation 6 is given by o= Ea Vu Following a method similar (0 Unt followed above, we have: (Eqn. 3.9) 2 Foe a oa7eoxy2 0.6745 From eqns. (3.22) and (3.24), we have probable error, r= 08453d r= 067450 Also, Standard devitation of mean and, Standard deviation of standard deviation, ee oeeenn Tables Bie pobebilty of an individual measurement to fall within the deviations x, 2 te %* fan and x, from the mean is givenBEd] > STATISTICAL ANALYSIS OF TEST DATA Pin) = Je nacre a for aNormal Gaussian curve ‘the parameter = —- and he mean value 7 may beknown from the raw data, Since the evalu theimegral ica cube anda tedious task, hence probability tables have es fn value of probability can be directly read off y= vie =+, Let Yih = 5 o, x= oy Now, Pawn) = Poy where, y, = v2hn eas Poni gn. (3.29) may be written as: 2 1 3 POpy (PF dy ~ Af IP ay Each of the two integrals in equation (3.30) is a function of a single parameter y, or yz To“ probability, = gle parameter 3, OF Ys * given in Ts limits.x, and x, Tabulated values of the int Aste cure is symmetrical he 3:1 sinc the vale of pony aoe alfeed ae Semiramis. yet o ‘The able3.1 indicatesthe probability value p= curve, an equal probability would exis forthe deviant deviation lies between + & works Out lo be 2 (3413. same table may be used for ve PO.y) = le 12) dy any in Table 3.1 and can be used to compute the proability be fory=Dand fey acshoen ey he imegal represent the area under the normalised Gaussian ‘substitution, deviation o, find the probability that the deviation lies between + 0. For ®* o 13 fory 1 ie between 0.6826, = 1.0. Due to symmetrical nature of ‘deviations replacing y by -» in Gand 0, Hence the prot” Pe | 06 10 MM 12 13 is 1s 16 Le 19. 21 23 24 | 2s 26 27 zs 29 30resents the interval about th Tor which te" the mean Which hy | jred to find the probable ©! eal area about the mat a = Pontere tt er ‘this ‘corresponds (0: sym ecing)Y ‘we look for the value of,*/ [Observation sume wc ete dat 05. Only Js t0 0.25. > ae i - probability table that comespoe ie “aoe 7 the probable error is given PY n=i0 (ii) The stands (iii) Probabl Ure ee While measuring a temperature the following ten readings were recorded : 49.6, 39.9, 39.7, 3.9, 40.0, 39.8, 39.9, 39.8, 40:4 and 39.7°C, Caleulate the following : ()/ The mean, 6) Rang ‘(ly/The standard deviaton, - ~{fii) The probable error of one reading. (jv) Probab \ (iv) The probable error of mean, and (0) The range, In atest tempera rm isi : Procedures. Aft EEIMIETM The satistical valves of the given data are computed with the help of the followin Temperature“ : Frequency of O Calculate 0 (i) arith (it) Averog (iit) Standa (iv) Varia ts) Proba (vi) Proba (vii) Standbeers ation wenaber Tempera core reading, °C we Devtation, The nican Value of t q = 7 87 i ant (i) The standard deviation a since the number Oat = Vion “TC (iii) Probable error of one readin ry = 0.6745 = 0.6745 x02 (i) Probable error of mean: (») Range: Range = 40.4 -39.6=0.8°C (Ans) tha test temperanare, measurements of Emer were made 100 timex with ¥ procedures. After applying the correction! Ine following results were ob 197 198 199 20 pa 2D 36 Temperature °C Frequency of Oveurrence Calculate the following (i) Arithmetic mean. rion. (il) Average dev (ili) Standard deviation. tiv} Variance. (v) Probable error of me readings (ot) Probable error of (vit) Standard deviation of ve mean, and, the deviation.