Lecture
Metrology — is the science of measurements, of the methods and means of ensuring their uniformity, and of the ways of achieving the required accuracy. The word “metrology” comes from the Greek words “metron” — measure, and “logos” — study.
Metrological support (MS) — is the establishment and application of the scientific and organizational foundations, technical means, rules and norms necessary to achieve uniformity and the required accuracy of measurements.
The scientific foundation of MS is metrology. The organizational foundation of MS is the metrological service of the Russian Federation, made up of the state and departmental metrological services, based on the fundamental provisions of legal metrology. The regulatory and legal foundation of MS is formed by the set of rules, requirements and norms established in the standards and normative documents on standardization in the Russian Federation.
The basis of the technical base of MS is measuring and monitoring instrumentation. The technical base of MS rests on the measurement-standard base of the Russian Federation, which consists of more than 150 state primary and special measurement standards and 60 secondary (working) measurement standards, providing for the maintenance and reproduction of 70 physical quantities in linear-angular, mechanical, temperature, thermophysical, electrical, magnetic, radio-engineering, optical and other kinds of measurements, across various amplitude, frequency and dynamic ranges. The ultimate goal of MS is to reduce to a rational minimum the possibility of erroneous decisions being made from the results of measurements, tests and monitoring.
The main problems studied in metrology are:
The measurement procedure generally consists of the following stages: adopting a model of the object of measurement, choosing the measurement method and the measuring instrument, and carrying out the experiment to obtain a result. All of these components lead to the result of measurement differing from the true value of the measurand.
Metrological characteristics. Characteristics that affect the results and errors of measurements are called metrological characteristics. The accuracy of results depends on how precisely they are maintained during manufacture and how stable they remain during operation. They include the transfer function (static characteristic of conversion), the sensitivity of the measuring instrument, the scale interval, the discrimination threshold, and also the dynamic characteristics.
Transfer function (static characteristic of conversion) — is the functional relationship between the informative parameters of the output and input signals of a measuring instrument. The transfer function adopted for a measuring instrument (type) and established in the scientific and technical documentation for that instrument (type) is called the nominal transfer function of the instrument (type). The nominal static conversion characteristic makes it possible to calculate the value of the input quantity from the value of the output quantity. It may be specified analytically, in tabular form, or graphically.
Sensitivity of a measuring instrument — is the ratio of the increment of the output signal Δy of the instrument to the change of the input signal Δx that caused this increment. In the general case, the sensitivity is
S = lim Δy/Δx = dy/dx.
With a nonlinear static conversion characteristic, the sensitivity depends on X, whereas with a linear characteristic it is constant. In measuring instruments with constant sensitivity, the scale is uniform, i.e., the distance between scale divisions is the same.
Scale interval — is the difference between the values of the quantities corresponding to two adjacent scale marks.
In instruments with a uniform scale the division value is constant; in instruments with a nonuniform scale it may differ across different parts of the scale, in which case the minimum division value is standardized. The division value of an instrument's scale can be determined from its absolute sensitivity and equals the number of units of the measured quantity per one scale division of the instrument (instrument constant): C =1/S.
Discrimination threshold — is the smallest change in the input quantity detectable by means of the given measuring instrument. The discrimination threshold is expressed in units of the input quantity.
The metrological characteristics also include dynamic characteristics, i.e., characteristics of the inertial properties (elements) of a measuring device that determine the dependence of a measuring instrument's output signal on quantities varying in time: parameters of the input signal, external influence quantities, and load. The dynamic characteristics include the transient response, the impulse transient response, the amplitude-phase characteristic, the transfer function, and others.
The dynamic properties of a measuring instrument determine its dynamic error.
Dynamic error — is the difference between the error of an instrument in dynamic mode and its static error.
Standardized metrological characteristics. For each type of instrument, based on its specifics and purpose, a certain set of metrological characteristics is standardized and specified in the normative-technical documentation. The general list of standardized metrological characteristics, the forms of their presentation, and the methods of standardization are established in the GOST. It includes:
Measurement errors. Measurement error — is the deviation of the measurement result from the true value of the measurand. Depending on the way the error is expressed, measurement errors are divided into absolute and relative.
Absolute error of measurement — is the difference between the measured value Xm of the physical quantity and its true value Xt, expressed in units of the measurand:
Δ = Xm – Xt.(1.15)
Relative error of measurement — is the ratio of the absolute error of measurement to the true value of the measurand (in %):
rel= (Δ/X)100%. (1.16)
In practice, instead of the true value of the measurand, the actual value Xa, obtained by means of a reference measuring instrument, is used. Expressions (1.15) and (1.16) then take the form
. (1-17)
The absolute error of measurement Δ, determined by expressions (1.15) and (1.17), is the total error made up of two components — systematic and random, i.e. Δ = Θ + δ.
Systematic error. Systematic error — is the component of the measurement error that remains constant or varies in a regular manner during repeated measurements of the same quantity. By the character of their manifestation, systematic errors are divided into constant and variable. Variable errors, in turn, may be progressive, periodic, or varying according to a complex pattern.
Constant systematic errors are those that remain unchanged throughout an entire series of measurements, for example, the error due to inexact adjustment of a reference measure, or the error due to inexact zero-setting of an instrument's pointer, etc.
Variable systematic errors change in the course of measurement. If during measurement the error monotonically decreases or increases, it is called progressive. For example, the error due to the discharge of an instrument's power supply changes monotonically if the measurement result depends on the supply voltage.
Periodic systematic error — is an error whose value is a periodic function of time. An example of it may be the error caused by daily variations in the supply voltage of the electrical mains. A systematic error may also vary according to some complex pattern. Such, for example, are the errors caused by inaccuracy in the application of an instrument's scale, the error of an electricity meter at different load values, the error caused by variations in ambient temperature, and others.
The nature and origin of systematic errors are usually determined by the specifics of the particular experiment. By their cause of occurrence, they can be divided into four main groups: instrumental, methodical, installation, and subjective.
Instrumental errors depend on the errors of the measuring instruments used. Inaccuracy of calibration, design imperfections, and changes in an instrument's characteristics during operation are the causes of instrumental errors. These are in turn divided into intrinsic and complementary errors.Intrinsic error of a measuring instrument is the error under conditions accepted as normal, i.e., at normal values of all quantities affecting the measurement result (temperature, humidity, supply voltage, etc.).Complementary error of a measuring instrument is the error additionally arising when the values of influencing quantities deviate from normal. Individual components of the complementary error are usually distinguished, for example the temperature error, the error due to changes in supply voltage, etc. The elimination of complementary errors has its own particular features.
Methodical errors arise from imperfections in the measurement method, from the use of simplifying assumptions and approximations in deriving the formulas applied, and from the influence of the measuring instrument on the object of measurement. For example, measuring temperature with a thermocouple may involve a methodical error caused by disturbance of the temperature regime of the object under study as a result of introducing the thermocouple into the measurement zone.
Installation errors are caused by improper use of a measure or instrument, or by deviation of external conditions from normal. For example, tilted installation of an instrument, the presence of an external magnetic field, deviation of temperature from normal, and others.
Subjective errors arise as a result of the particular features of the observer. This may happen, for example, because of an incorrect direction of gaze when observing the readings of a pointer instrument (parallax error), or because of the observer's tendency to over- or underestimate the results, and others. The use of digital instruments and automatic measurement methods makes it possible to eliminate errors of this kind.
A systematic error is called inherent if its occurrence is due only to the essence of the measurement method or the formula by which the result is calculated, and other such causes, and does not depend on the quality of manufacture or the conditions of use of the measuring instrument.
Systematic errors remain constant during repeated measurements or vary according to a definite pattern, and do not depend on the number of measurements. The distortions they introduce into the measurement result can be eliminated or accounted for. An unexcluded or undetected systematic error is more dangerous than a random one. Whereas random errors determine the reliability of a result, systematic errors persistently distort it.
Methods of eliminating systematic errors. Systematic errors can in principle be identified and excluded from measurement results by introducing corrections, eliminating the sources of error themselves, and by the methods of double measurement and substitution.
Before beginning an experiment, the causes giving rise to systematic errors should be eliminated wherever possible. Instrumental errors of measures and instruments are accounted for by introducing corrections. A correction is the value of a quantity, of the same kind as the measured one, that must be added to the value obtained by measurement in order to exclude a systematic error. Introducing corrections is the most widely used method of excluding systematic instrumental errors. A correction is determined by means of verification of technical equipment and by compiling and using appropriate tables and graphs. Computational methods of finding correction values are also used.
Installation error is eliminated by observing the requirements for the operation of the measuring instrument. Subjective error is reduced and, wherever possible, eliminated by using qualified specialists to carry out the measurements and various methods of checking the results of these measurements.
The method of compensating the error by sign is used to exclude systematic errors which, depending on the measurement conditions, may enter the measurement result with one sign or the other, for example errors from thermo-EMF, or from the influence of the intensity of a constant electric or magnetic field. In this case the measurement should be carried out twice, so that the error enters the measurement results once with one sign and the other time with the opposite sign. The average of the results of these two measurements will be free of systematic error. In automatic measurements, circuit methods of correcting systematic errors are widely used, for example the compensating connection of transducers, various temperature- and frequency-correction circuits, and others.
The substitution method consists in the measured quantity being replaced by a known quantity obtained by means of an adjustable measure. If such a substitution is carried out without any other changes in the experimental setup, and after substitution the same instrument readings are established, then the measured quantity is equal to the known quantity, the value of which is read off the indicator of the adjustable measure. This technique makes it possible to exclude constant systematic errors. The error of measurement when using the substitution method is determined by the error of the measure and by the error arising when reading off the value of the quantity that replaces the unknown one.
It should be noted that the exclusion of systematic errors by the methods indicated above is carried out down to the level of unexcluded systematic errors, an estimate of whose total component is found on the basis of information about the metrological characteristics of the technical means used. If such information is insufficient, it may be useful to compare the measured values with similar results obtained in other laboratories.
The use of microprocessor devices in measuring instruments makes it possible to almost completely exclude or correct many kinds of the systematic component of error, especially instrumental errors. The automatic introduction of corrections related to calibration inaccuracies, the calculation and exclusion of complementary errors, and the correction of the additive and multiplicative components of measurement error make it possible to substantially improve measurement accuracy.
Random error. The random component of the error changes randomly during repeated measurements of the same quantity. It is usually the result of the simultaneous action of many independent causes, each of which individually has little effect on the measurement result. Random errors cannot be excluded from the measurement result, but probability theory and mathematical statistics make it possible to assess the measurement result in the presence of random errors. They are characterized by properties expressed in two axioms:
1. The axiom of randomness — with a very large number of measurements, random errors equal in magnitude and different in sign occur equally often. The number of negative errors equals the number of positive errors.
2. The axiom of distribution — small errors occur more often than large ones. Very large errors do not occur. Accepting these two axioms allows random errors to be treated as random variables obeying some symmetric distribution law. In assessing the accuracy of the result obtained, the form of the distribution law of the random errors must be taken into account. In the practice of electrical measurements, various distribution laws of random errors are encountered: the uniform symmetric distribution law (rounding, reading, and quantization errors), the normal distribution law (errors from thermal noise, the total error of a large number of components), the bimodal, and the triangular (Simpson's law) distribution, and others.
Determination of the confidence limits Δcof the random component of the error of measurement result Δ is carried out on the basis of the computed estimate of the standard deviation —σ(X) taking into account the specified confidence levelPconf and the number of observationsn. Assuming a normal distribution of the random variableX with a limited number of measurements (fewer than 30) and a specified confidence levelPconf, the confidence limits of the random component of the result's error are determined taking into account Student's correction coefficientt(n):
± Δr= ±t(n)σ(X).
With a large number of measurements (> 30) and a normal distribution of the random variable X, the probability of finding the error Pconf within the specified limits ± Δris equal to
Pconf(-Δr < Δ < Δr) = 2Φ(Δr/σ(X)),
where Φ(z) — is the tabulated integral of the Laplace function;z — is the argument of the Laplace function.
The confidence interval and confidence level are chosen depending on the specific conditions of the experiment. Depending on the value of the measurand, the absolute error of measurement Δ, determined by expression (1.15) and (1.17), is also a total error made up of two components: an additive component, whose values do not depend on the value of the measurand X, and a multiplicative component, whose values depend on the value ofX, i.e.
Δ = Δadd+ Δm. (1.18)
A measurement result is fit for further use only when, in addition to the measured value of the physical quantity, the value of the error is also indicated. The error of the result of a direct single measurement depends on many factors, but it is primarily determined by the error of the measuring instruments used. Therefore, to a first approximation, the error of the measurement result can be taken equal to the error that characterizes the measuring instrument used at the given point on its scale. Both the absolute and the relative errors of the measurement result must be calculated, since the first is needed for rounding the result and recording it correctly, while the second is needed for an unambiguous comparative characterization of its accuracy.
The results of multiple observations obtained in direct measurements of a physical quantity are called equally accurate (equally scattered), if they are independent, identically distributed random variables. In this case the measurements are carried out by a single observer under the same environmental conditions and using the same measuring instrument. An accurate estimate of the actual value of the measurand under equally accurate measurements can be obtained only by statistical processing of a group of measurement results.
Forms of presenting measurement results. The final measurement result is presented in one of four forms:
1) by an interval within which, at a specified probability, the total error of measurement lies;
2) by an interval within which, at a specified probability, the systematic component of the error lies, together with a standard approximation of the distribution function of the random component of the measurement error and the standard deviation of the random component of the measurement error;
3) by standard approximations of the distribution function of the systematic and random components of the measurement error and their standard deviations;
4) by the distribution functions of the systematic and random components of the measurement error.
The choice of the form of presenting the measurement result is determined by the purpose of the measurements and the way their results are used.
Unequally accurate measurements. In measurement practice there also occur unequally accurate measurements, when measurements of the same physical quantity are carried out by several observers of different qualification and experience, on instruments of different accuracy classes, or over the course of several days. The obtained arithmetic-mean values of the individual samples differ from one another; therefore, in assessing the measurement result and its error, the degree of confidence in the obtained sample means is taken into account in the form of a “weight,” which is assigned to each series of measurements in proportion to one of the parameters (the probability, the number of measurements, the value of the standard deviation), or by the method of expert assessment. The greater the degree of confidence in a result, the greater the number expressing its weight. If it is established that all the samples of unequally accurate measurements belong to a single general population, the statistical parameters of that population are determined and the limits of the confidence probability are established from Student's distribution.
The value of the measured quantity closest to the true value is:
X' — -
where T], X2, ...,X„ — are the mean values for individual groups of measurements;p', p 2, -, — are their weights;xq — is the weighted mean value of the measurand.
The basis of the calculation is usually the standard errors. The weights of the corresponding groups of measurements are taken to be inversely proportional to their variances, i.e., the relationship used is P\ : P\ : P\ :P\ = 1/σ2, : 1/σ22: 1/σ23: 1/σ24.
The standard error of the weighted-mean value S0 is determined by the formula
where Pi — is the weight of each measurement result;m — is the number of series of measurements.
Indirect measurements. These are measurements in which the sought value of the physical quantity Q is found on the basis of a known relationshipQ = f(x, y, z) between this quantity and the quantitiesx, y, z, that are subjected to direct measurement. For example, measuring powerP =UI from the measured values of currentIand voltageU.
To obtain an estimate of the systematic error of the result of an indirect measurement, using the expansion of the functionQ = f(x, y, z) in a Taylor series and limiting it to its linear part, one obtains
(1.19)
∂x
∂y
The quantities ∂f / ∂x, ∂f /∂y, ∂f / ∂z are called the partial derivatives of the indirect measurement.
Random error of an indirect measurement
(1.20)
where σQ — is the standard deviation of the result of the indirect measurement.
Law of error summation. The problem of summing errors arises in the analysis both of individual measuring transducers and of a measuring device as a whole. If a measuring device is a chain of measuring transducers, the total number of components of its error can reach 10...50 or more.
The only method of isolating the systematic component of the error of an instrument is the method of statistical testing, i.e., carrying out multiple repeated verifications of the instrument. If, in doing so, an error of a definite sign and magnitude is consistently observed across a whole series of measurements, it can be classified as systematic and excluded from probabilistic consideration. Equipment that has just been manufactured and has not yet undergone adjustment may have arbitrarily large systematic errors. During fitting and calibration these errors are eliminated as far as possible, and then a process of their gradual elimination proceeds, which in the general case is random in nature. The process of eliminating errors after the moment of verification can develop in two directions:
The determining criterion in choosing the method of error summation is not the division of errors into systematic and random, but the criterion of their strong or weak mutual correlation. For example, a moving-coil (magnetoelectric) measuring mechanism, when the temperature changes, has a positive error from the decrease in spring stiffness and a negative error from the decrease in magnet induction. Under random temperature fluctuations, both of these error components manifest themselves as random. However, despite the random character of their occurrence in time, they are rigidly linked (strongly correlated) with each other, since under any random fluctuations a positive value of one of them is always accompanied by a negative value of the other. Therefore these errors must always be subtracted from each other.
Probability theory gives the following expression for the variance of the sum of two random variables:
σ² = σ1² + σ2² + 2rσ1σ2
where r — is the correlation coefficient of these quantities.
For the case of strongly correlated random variables r ≈ ±1 we obtain algebraic summation of the components taking into account their sign
σΣ=σ1+σ2. (1-21)
With weak correlation or its absence (r ≈ 0) we obtain geometric summation of the components.
In determining the total error of a device, a simplified approach to determining the mutual correlation of errors is used. If a number of errors of one or several transducers are caused by one and the same cause, as a result of which they turn out to be fairly strongly correlated, then their mutual correlation coefficient is taken equal to ± 1.
+If, on the other hand, the errors are caused by reasons that have no obvious connection with one another, their correlation is taken as zero. No intermediate values are used. Proceeding from this, in order to sum errors it is first necessary to identify groups of errors that are strongly correlated with one another. Owing to the rigid mutual correlation and the common cause producing these errors, they will be distributed according to the same law, and the shape of the resulting distribution law will correspond to that same law. Therefore, within each of these groups the errors must be added algebraically, taking their signs into account. The resulting errors obtained after summation within each group no longer have rigid correlational links between them and should be regarded as statistically independent.
Uniformity of measurements — a state of measurements in which their results are expressed in legalized units and the errors of the measurement results are known with a given probability. It makes it possible to compare measurements performed at different times, by different means and methods. Uniformity of measurements is ensured by the uniformity of measuring instruments and the correctness of the procedures used to perform them. Here, byuniformity of measuring instruments is meant a state in which they are calibrated in legalized units and their metrological properties conform to established norms.
A physical quantity (hereinafter, a quantity) denotes a property that is qualitatively common to many physical objects, physical systems, their states and the processes occurring in them, but quantitatively individual for each object. The entire variety of quantities is combined intosystems of physical quantities related to one another by dependences. A system includes base and derived quantities. Each system has its own set of these quantities. To designate a system of quantities, a group of base quantities is indicated.Base is the name given to a quantity that is included in a system and conventionally accepted as independent of the other quantities of that system.Derived — a quantity included in a system and defined through the base quantities of that system.The dimension of a quantity is the expression reflecting the relationship of a derived quantity to the base quantities of the system, representing a product of base quantities raised to the corresponding power, called thedimensional exponent of the quantity. A system of quantities may consist of both dimensional and dimensionless quantities.Dimensional is the name given to a quantity in whose dimension at least one of the base quantities is raised to a power not equal to zero,dimensionless — a quantity in whose dimension the base quantities appear to a power equal to zero. A dimensionless quantity in one system of quantities may be a dimensional quantity in another system.
For each quantity included in a system, a unit of the quantity is chosen and assigned a numerical value equal to unity. In accordance with the kinds of quantities, a distinction is made between base and derived units, which form, for a given system of quantities,a system of units. As base units one chooses those which, firstly, can be reproduced with the highest accuracy, and secondly, are convenient in measurement practice or in their reproduction. Units of quantities included in a system are calledsystem units. In addition to system units,non-system units are also used, not belonging to any of the systems of units: the unit of power — horsepower, units of time — hour, day, and so on. They arose in the course of the development of measurement technology to satisfy practical needs, or were introduced for convenience of use in measurements. Multiples and submultiples of units of quantities are used for the same purposes.A multiple unit is one that is a whole number of times larger than the system or non-system unit: kilohertz, megawatt;a submultiple unit — one that is a whole number of times smaller than the system or non-system unit: milliampere, microvolt. Strictly speaking, many non-system units can be regarded as multiple or submultiple units.
In the past there existed several versions of systems of units, such as, for example, CGS (centimetre, gram, second), MKS (metre, kilogram, second). The most widely used throughout the world is the International System of Units SI, which includes the MKS system of units (mechanical units) and the MKSA system (electrical units). The SI system contains seven base units and two supplementary units. Each base unit has a name, a symbol and its own definition:
Two supplementary units belonging to the system serve to measure plane and solid angles, named respectively the radian (rad) and thesteradian (sr). Supplementary units are not classified as derived units, since they do not depend on the base units. The International System SI is considered the most perfect and universal in comparison with the systems that preceded it. In addition to the base and supplementary units, the SI system also has a large number of derived units of space and time, mechanical quantities, electrical and magnetic quantities, thermal, light and acoustic quantities, as well as ionizing radiation. After the adoption of the International System of Units by the General Conference on Weights and Measures (CGPM), practically all the largest international organizations included it in their recommendations on metrology and called on all the member countries of these organizations to adopt it. In Russia the SI system was officially adopted in 1963 through the introduction of the corresponding state standard. Today the SI system has indeed become international, but at the same time non-system units are also used, for example the tonne, the day, the litre, the hectare, and others.
Depending on the limits of the permissible intrinsic and complementary error, all measuring instruments are divided into accuracy classes.
Accuracy class characterizes the accuracy of a measuring instrument, but it is not a direct indicator of the accuracy of measurements performed with the aid of that instrument. A special kind of measuring instrument is the measurement standard.A measurement standard — is a measuring instrument (or a complex of measuring instruments) of the highest accuracy, ensuring the reproduction and (or) storage of a unit of measurement for the purpose of transferring its size to measuring instruments lower down the verification hierarchy, made to a special specification and officially approved in the established manner as a standard. Standards are subdivided, by subordination, into primary, special and secondary.
Primary standards reproduce units with the highest accuracy possible in a given field of measurement at the current level of scientific and technical achievement. A primary standard may be national (state) or international. A national standard is approved as the initial measurement instrument for a country by the national metrology body. In Russia national (state) standards are approved by Gosstandart of the Russian Federation.
A special standard ensures the reproduction of a unit of a physical quantity under special conditions and replaces the primary standard for those conditions. International standards are kept and maintained by the International Bureau of Weights and Measures (BIPM). The most important task of the BIPM's work consists in the systematic international comparison of the national standards of the largest metrological laboratories of different countries, which is necessary to ensure the reliability, accuracy and uniformity of measurements as one of the conditions for international economic relations. Both standards of the base quantities of the SI system and of derived quantities are subject to comparison. Definite periods of comparison have been established for them. For example, the standards of the metre and the kilogram are compared every 25 years, while electrical and photometric standards are compared once every 3 years.
The values of secondary standards are established against the primary standard. Secondary standards are divided into working standards, witness standards, copy standards and transfer standards.
A working standard is intended for transferring the size of a unit to reference instruments of the highest accuracy,a witness standard — is used for checking the integrity of the state standard and for replacing it in case of damage or loss,a copy standard — is used for transferring the size of the unit to working standards, and it is not always a physical copy of the state standard.A transfer standard (travelling standard, transportable standard) is intended for the comparison of standards which, for various reasons, cannot be compared directly with one another, for example during international comparisons or when a standard needs to be transported, and so on. Secondary standards are approved by Gosstandart of the Russian Federation or by state scientific metrological centres, in connection with the particular features of their use. The creation and improvement of a base of standards is an extremely complex task, requiring great effort from teams of scientists, engineers, designers and production workers. This is explained by the high accuracy of the standards for electrical units, which exceeds the accuracy of the measuring instruments in use by tens of times. In modern standards the latest achievements of physics, chemistry and technology are being applied ever more widely. Work with standards is carried out in accordance with the rules for their storage and use, which specify the procedure for their use, monitoring of correct storage, and the conduct of studies and comparisons of standards, including international ones.
In recent years high results in the accuracy and reliability of standards have been obtained, created on the basis of quantum effects, which suggests the possibility of creating new standards. Quantum standards are characterized by a high degree of stability in the values of the reproduced units of quantities. Using quantum effects, a modern standard of the ampere and of the ohm has been created. If a standard of mass is also created on the basis of the possibilities of nuclear physics, then many existing standards will pass into the category of “eternal” ones, since the dimensions of their quantities are related to mass.
Standardization — is an activity aimed at the development and establishment of requirements, norms, rules and characteristics, recommended and mandatory for compliance, ensuring the consumer's right to purchase goods of appropriate quality, as well as the right to safety and comfort in labour.
The purpose of standardization — is to achieve the optimal degree of ordering in a given field of activity through the broad and repeated use of established provisions, requirements and norms for solving actually existing or potential problems. The principal result of standardization activity should be an increase in the degree of conformity of a product (service) and of processes to their functional purpose, the elimination of technical barriers in international trade, and the promotion of scientific and technical progress in various spheres of human activity.
Standardization is connected with such important concepts as the object of standardization and the field of standardization. The object of standardization — is a product, work (process) or service that is subject to, or has undergone, standardization.The field of standardization — is a set of interrelated objects of standardization.
In the process of standardization, norms, rules, requirements and characteristics relating to the object of standardization are developed, which are drawn up in the form of a normative document having a legislative or recommendatory character.A standard — is a normative document on standardization, developed, as a rule, on the basis of consensus, characterized by the absence of objections on substantive issues from the majority of interested parties, and adopted (approved) by a recognized body (enterprise).
In the Russian Federation, normative documents on standardization are divided into the following categories: state standards of the Russian Federation — GOST R; industry standards — OST; technical specifications — TU; standards of enterprises and associations of enterprises (unions, associations, concerns, joint-stock companies, cross-industry, regional and other associations) — STP; standards of scientific-technical and engineering societies (unions, associations and other public associations) — STO.
Normative documents on standardization also include the all-Russian classifiers of technical and economic information, the procedure for the development and application of which is established by Gosstandart of the Russian Federation.
Standardization work in Russia is carried out by a governing body (national), working bodies and supervisory organizations. The national standardization body in Russia is the Committee of the Russian Federation for Standardization, Metrology and Certification (Gosstandart of the Russian Federation). It forms and implements state policy in the field of standardization, exercises state control and supervision over compliance with the mandatory requirements of state standards, participates in international (regional) standardization work, organizes professional training and retraining of personnel in the field of standardization, and also establishes the rules for applying international (regional) standards, rules, norms and recommendations on standardization within the territory of the Russian Federation, unless otherwise established by international treaties of the Russian Federation.
The development of state standards of the Russian Federation is carried out, as a rule, by technical committees (TCs) for standardization in accordance with the assignments of the state standardization plans of the Russian Federation, the programmes (plans) of work of technical committees, and contracts for the development of standards. Technical committees are set up on the basis of enterprises (organizations) that specialize in particular kinds of products and technologies or kinds of activity and possess the highest scientific and technical potential in the given field. In developing standards, the developers are guided by the current legislation of the Russian Federation, the state standards of the State System of Standardization (GSS) of the Russian Federation, and other normative documents on standardization, and the documents of international and regional standardization organizations are also taken into account.
The leading role in providing information support for the work of the standardization bodies of all countries of the world is played by the International Organization for Standardization (ISO) — through the Committee on Information Systems and Services (INFCO), which is accountable to the ISO General Assembly. The Assembly determines the directions of its activity, its goals and objectives, and INFCO regularly reports to it on the work performed. ISO's information system — ISONET — is part of INFCO's information group. The priority goals of ISONET are: ensuring the exchange of information on international and national standards, on standardization documents (including governmental ones), and on publications of books, reference works and educational literature in the field of standardization; establishing contacts with the information systems of other international organizations (the UN, UNESCO, the IAEA, and others); and creating a unified information language, a thesaurus. Russia is represented in ISONET by Gosstandart of the Russian Federation.
The International Classifier for Standardization (ICS) plays a major role in providing information support, serving as the methodological basis for preparing national indexes of standards.
Standards that are part of the state system for ensuring the uniformity of measurements (GSI) regulate the basic rules, norms and provisions in the field of ensuring the uniformity of measurements, the procedure for the approval and development of standards for physical units, requirements for the procedures and schemes for the verification of measuring instruments, their state testing and certification, and provisions and requirements for systems of standard reference data and reference materials.
Industry systems for ensuring the uniformity of measurements (OSI), being an integral part of the GSI, establish the scientific, technical and organizational foundations and tasks of the metrological support of production, taking into account its particular features. The state and industry systems for ensuring the uniformity of measurements regulate the choice of measuring instruments for monitoring production processes and carrying out measurements.
GSI standards also establish requirements for the structure, content and presentation of normative and technical documents on measurement procedures.
A number of organizations operate in the field of international standardization, the most representative of which are: the International Organization for Standardization (ISO), the International Electrotechnical Commission (IEC), the European Organization for Quality (EOQ), the International Organization of Weights and Measures (IOWM), the International Organization of Legal Metrology (OIML), the European Economic Community (EEC), the European Committee for Standardization (CEN), and others.
The ISO/IEC system is the largest of the existing international technical organizations, extending its activity to all branches of the economy and science. The principal goal of ISO/IEC is to ensure the development of standardization and related fields in order to promote the international exchange of goods and services, as well as the development of cooperation in intellectual, scientific-technical and economic activity.
The oldest intergovernmental scientific-technical organization, the IOWM, was founded on 20 May 1875 in accordance with the Metre Convention, signed by 17 countries (including Russia), with the aim of unifying the systems of units of measurement used in different countries and establishing actual uniformity of the standards of length and mass (the metre and the kilogram).
Certification — is a procedure by which assurance is given that a product, process or service conforms to specified requirements. The result of certification is a document called acertificate of conformity.
The establishment of a product's conformity to specified requirements is carried out in testing laboratories by means of tests. By a test is meant a technical operation consisting in the determination of one or several characteristics of a given product in accordance with an established procedure, according to accepted rules.
A systematic verification of the degree of conformity to specified requirements is called conformity assessment. More often the termcontrol is used, which is regarded as an assessment of conformity by measuring specific characteristics of the product.
Verification, supervision and assurance of conformity are related to conformity assessment. Verification of conformity — is confirmation of the conformity of a product (process, service) to established requirements through the examination of evidence.Supervision of conformity — is a repeated assessment carried out to ensure that a product (process, service) continues to conform to established requirements.Assurance of conformity — is a product-testing procedure whose result is a document calleda manufacturer's declaration of conformity (declaration) — a written guarantee that the manufacturer's product conforms to the specific requirements of a normative document. It contains reference information about the manufacturer and the product.
Confirmation of conformity through certification presupposes the mandatory participation of a third party. A third party — is a system for certifying homogeneous products, independent of the supplier (the first party) and of the purchaser (the second party).
Certification systems make use of the services of testing laboratories, which may be independent organizations or an integral part of a certification body or another organization.
Accreditation — is the official recognition of a testing laboratory's right to carry out specific tests or specific types of tests. Accreditation is preceded bycertification — an examination of a testing laboratory to establish its conformity to the accreditation criteria. It represents an assessment of the state of affairs in the laboratory according to certain parameters and criteria, the choice of which is based on the general requirements for testing laboratories discussed above.
Laboratory accreditation — is an independent field of activity, linked with certification. There exist variousaccreditation systems, having their own rules of procedure and management. An accreditation system is managed by an accreditation body, which may itself carry out the accreditation of testing laboratories, and may also delegate, in full or in part, its powers of certification to a certification agency or another competent organization. In third-party certification systems, two methods are used for indicating conformity to standards: the certificate of conformity and the mark of conformity, which are the means of informing all interested parties about a certified product.
A certificate of conformity — is a document, issued according to the rules of the certification system, stating that the necessary assurance is provided that a duly identified product (process, service) conforms to a specific standard or other normative document. The certificate may relate to all the requirements of the standard, or to individual sections or specific characteristics of the product, which is clearly stipulated in the document itself. The information presented in the certificate must make it possible to compare it with the results of the tests on the basis of which it was issued.
A mark of conformity — is a mark protected in the established manner, applied (or issued by a certification body) in accordance with the rules of the certification system, indicating that the given product (process, service) conforms to a specific standard. Permission (a licence) to use the mark of conformity is issued by the certification body.
Certification may be either mandatory or voluntary in character. Mandatory certification is carried out on the basis of laws and legislative provisions and provides evidence of a product's (process's, service's) conformity to the requirements of technical regulations and the mandatory requirements of standards. Since the mandatory requirements of these normative documents relate to safety and the protection of human health and the environment, the principal aspects of mandatory certification are safety and environmental compatibility. The list of objects subject to mandatory certification is established at the state level of government.
Voluntary certification is conducted on the initiative of legal entities or individuals, on contractual terms between the applicant and the certification body, withinvoluntary certification systems. Voluntary certification may also be conducted, in mandatory certification systems, by mandatory certification bodies. The normative document against which conformity is tested in voluntary certification is chosen, as a rule, by the applicant. The applicant may be a manufacturer, supplier, seller or consumer of the product. Voluntary certification systems most often bring together manufacturers and consumers of products who are interested in developing trade on the basis of long-term partnership relations.
+Unlike mandatory certification, whose objects and the confirmation of their conformity are linked to legislation, voluntary certification concerns kinds of products (processes, services) that are not included in the mandatory list and are determined by the applicant. The rules and procedures of a voluntary certification system are determined by the voluntary certification body. However, just as in mandatory certification systems, they are based on the recommendations of international and regional organizations in this field. A decision in favour of voluntary certification is usually connected with problems of product competitiveness, the promotion of products on the market (especially abroad), and the preferences of purchasers, who are increasingly guided in their choice by certified products. As a rule, the development of voluntary certification is supported by the state.
Comments