Geometrical Product Specifications (GPS) - Inspection by measurement of workpieces and measuring equipment - Part 2: Guide to the estimation of uncertainty in GPS measurement, in calibration of measuring equipment and in product verification

Spécification géométrique des produits (GPS) — Vérification par la mesure des pièces et des équipements de mesure — Partie 2: Guide pour l'estimation de l'incertitude dans les mesures GPS, dans l'étalonnage des équipements de mesure et dans la vérification des produits

Specifikacija geometrijskih veličin izdelka - Preiskave z merjenjem izdelka in merilna oprema - 2. del: Navodila za ocenjevanje negotovosti pri meritvah geometrijskih veličin izdelkov, pri umerjanju merilne opreme in pri potrjevanju proizvodov

General Information

Status
Withdrawn
Publication Date
15-Dec-1999
Withdrawal Date
15-Dec-1999
Current Stage
9599 - Withdrawal of International Standard
Start Date
12-Apr-2011
Completion Date
13-Dec-2025

Relations

Technical specification
ISO/TS 14253-2:1999 - Geometrical Product Specifications (GPS) -- Inspection by measurement of workpieces and measuring equipment
English language
73 pages
sale 15% off
Preview
sale 15% off
Preview
Technical specification
ISO/TS 14253-2:2002
English language
73 pages
sale 10% off
Preview
sale 10% off
Preview
e-Library read for
1 day
Technical specification
ISO/TS 14253-2:1999 - Spécification géométrique des produits (GPS) -- Vérification par la mesure des pieces et des équipements de mesure
French language
74 pages
sale 15% off
Preview
sale 15% off
Preview

Frequently Asked Questions

ISO/TS 14253-2:1999 is a technical specification published by the International Organization for Standardization (ISO). Its full title is "Geometrical Product Specifications (GPS) - Inspection by measurement of workpieces and measuring equipment - Part 2: Guide to the estimation of uncertainty in GPS measurement, in calibration of measuring equipment and in product verification". This standard covers: Geometrical Product Specifications (GPS) - Inspection by measurement of workpieces and measuring equipment - Part 2: Guide to the estimation of uncertainty in GPS measurement, in calibration of measuring equipment and in product verification

Geometrical Product Specifications (GPS) - Inspection by measurement of workpieces and measuring equipment - Part 2: Guide to the estimation of uncertainty in GPS measurement, in calibration of measuring equipment and in product verification

ISO/TS 14253-2:1999 is classified under the following ICS (International Classification for Standards) categories: 17.040.01 - Linear and angular measurements in general; 17.040.40 - Geometrical Product Specification (GPS). The ICS classification helps identify the subject area and facilitates finding related standards.

ISO/TS 14253-2:1999 has the following relationships with other standards: It is inter standard links to ISO 14253-2:2011. Understanding these relationships helps ensure you are using the most current and applicable version of the standard.

You can purchase ISO/TS 14253-2:1999 directly from iTeh Standards. The document is available in PDF format and is delivered instantly after payment. Add the standard to your cart and complete the secure checkout process. iTeh Standards is an authorized distributor of ISO standards.

Standards Content (Sample)


TECHNICAL ISO/TS
SPECIFICATION 14253-2
First edition
1999-12-01
Geometrical Product Specifications
(GPS) — Inspection by measurement of
workpieces and measuring equipment —
Part 2:
Guide to the estimation of uncertainty
in GPS measurement, in calibration
of measuring equipment and in product
verification
Spécification géométrique des produits (GPS) — Vérification par la mesure
des pièces et des équipements de mesure —
Partie 2: Guide pour l'estimation de l'incertitude dans les mesures GPS,
dans l'étalonnage des équipements de mesure et dans la vérification
des produits
Reference number
©
ISO 1999
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but shall not
be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In downloading this
file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat accepts no liability in this
area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation parameters
were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In the unlikely event
that a problem relating to it is found, please inform the Central Secretariat at the address given below.
© ISO 1999
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means, electronic
or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or ISO's member body
in the country of the requester.
ISO copyright office
Case postale 56 � CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 734 10 79
E-mail copyright@iso.ch
Web www.iso.ch
Printed in Switzerland
ii © ISO 1999 – All rights reserved

Contents Page
Foreword.iv
Introduction.v
1 Scope .1
2 Normative references .2
3 Terms and definitions .2
4 Symbols .6
5 Concept of the iterative GUM-method for estimation of uncertainty of measurement .7
6 Procedure for Uncertainty MAnagement — PUMA .8
7 Sources of errors and uncertainty of measurement.13
8 Tools for the estimation of uncertainty components, standard uncertainty and expanded
uncertainty.17
9 Practical estimation of uncertainty — Uncertainty budgeting with PUMA.26
10 Applications .30
Annex A (informative) Example of uncertainty budgets — Calibration of a setting ring.34
Annex B (informative) Example of uncertainty budgets — Design of a calibration hierarchy.41
Annex C (informative) Example of uncertainty budgets — Measurement of roundness .65
Annex D (informative) Relation to the GPS matrix model.71
Bibliography.73
© ISO 1999 – All rights reserved iii

Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies (ISO
member bodies). The work of preparing International Standards is normally carried out through ISO technical
committees. Each member body interested in a subject for which a technical committee has been established has
the right to be represented on that committee. International organizations, governmental and non-governmental, in
liaison with ISO, also take part in the work. ISO collaborates closely with the International Electrotechnical
Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 3.
The main task of technical committees is to prepare International Standards. Draft International Standards adopted
by the technical committees are circulated to the member bodies for voting. Publication as an International
Standard requires approval by at least 75 % of the member bodies casting a vote.
In other circumstances, particularly when there is an urgent market requirement for such documents, a technical
committee may decide to publish other types of normative document:
— an ISO Publicly Available Specification (ISO/PAS) represents an agreement between technical experts in an
ISO working group and is accepted for publication if it is approved by more than 50 % of the members of the
parent committee casting a vote;
— an ISO Technical Specification (ISO/TS) represents an agreement between the members of a technical
committee and is accepted for publication if it is approved by 2/3 of the members of the committee casting a
vote.
An ISO/PAS or ISO/TS is reviewed every three years with a view to deciding whether it can be transformed into an
International Standard.
Attention is drawn to the possibility that some of the elements of this Technical Specification may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO/TS 14253-2 was prepared by Technical Committee ISO/TC 213, Dimensional and geometrical product
specifications and verification.
ISO 14253 consists of the following parts, under the general title Geometrical product specifications (GPS) —
Inspection by measurement of workpieces and measuring equipment:
� Part 1: Decision rules for proving conformance or non-conformance with specification
� Part 2: Guide to the estimation of uncertainty in GPS measurement, in calibration of measuring equipment and
in product verification [Technical Specification]
� Part 3: Procedures for evaluating the integrity of uncertainty in measurement values
Annexes A to D of this Technical Specification are for information only.
iv © ISO 1999 – All rights reserved

Introduction
This Technical Specification is a global GPS technical report (see ISO/TR 14638:1995). This global GPS Technical
Report influences chain link 4, 5 and 6 in all chains of standards.
For more detailed information of the relation of this report to other standards and the GPS matrix model, see
annex D.
This Technical Specification is developed to support ISO 14253-1. This Technical Specification establishes a
simplified, iterative procedure of the concept and the way to evaluate and determine uncertainty (standard
uncertainty and expanded uncertainty) of measurement, and the recommendations of the format to document and
report the uncertainty of measurement information as given in "Guide to the expression of uncertainty in
measurement" (GUM). In most cases only very limited resources are necessary to estimate uncertainty of
measurement by this simplified, iterative procedure, but the procedure may lead to a slight overestimation of the
uncertainty of measurement. If a more accurate estimation of the uncertainty of measurement is needed, the more
elaborated procedures of the GUM must be applied.
This simplified, iterative procedure of the GUM methods is intended for GPS measurements, but may be used in
other areas of industrial (applied) metrology.
Uncertainty of measurement and the concept of handling uncertainty of measurement being of importance to all the
technical functions in a company, this Technical Specification relates to e.g. management function, design and
development function, manufacture function, quality assurance function, metrology function, etc.
This Technical Specification is of special importance in relation to ISO 9000 quality assurance systems, where
it is a requirement that the uncertainty of measurement is known [e.g. 4.11.1, 4.11.2 a) and 4.11.2 b) of
ISO 9001:1994].
In this Technical Specification the uncertainty of the result of a process of calibration and a process of
measurement is handled in the same way:
� calibration is treated as "measurement of metrological characteristics of a measuring equipment or a
measurement standard";
� measurement is treated as "measurement of geometrical characteristics of a workpiece".
Therefore, in most cases no distinction is made in the text between measurement and calibration. The term
"measurement" is used as a synonym for both.
© ISO 1999 – All rights reserved v

TECHNICAL SPECIFICATION ISO/TS 14253-2:1999(E)
Geometrical product specifications (GPS) — Inspection by
measurement of workpieces and measuring equipment —
Part 2:
Guide to the estimation of uncertainty in GPS measurement, in
calibration of measuring equipment and in product verification
1 Scope
This Technical Specification gives guidance on the implementation of the concept of "Guide to the estimation of
uncertainty in measurement" (in short GUM) to be applied in industry for the calibration of (measurement)
standards and measuring equipment in the field of GPS and the measurement of workpiece GPS-characteristics.
The aim is to promote full information on how to achieve uncertainty statements and provide the basis for
international comparison of results of measurements and their uncertainties (relationship between purchaser and
supplier).
This Technical Specification is intended to support ISO 14253-1. This Technical Specification and ISO 14253-1 are
beneficial to all technical functions in a company in the interpretation of GPS specifications (i.e. tolerances of
workpiece characteristics and values of maximum permissible errors (MPE) for metrological characteristics of
measuring equipment).
This Technical Specification introduces the Procedure for Uncertainty MAnagement (PUMA), which is a practical,
iterative procedure based on the GUM for estimating uncertainty of measurement without changing the basic
concepts of the GUM and is intended to be used generally for estimating uncertainty of measurement and giving
statements of uncertainty for:
� single results of measurement;
� comparison of two or more results of measurement;
� comparison of results of measurement — from one or more workpieces or pieces of measurement equipment
— with given specifications [i.e. maximum permissible errors (MPE) for a metrological characteristic of a
measurement instrument or measurement standard, and tolerance limits for a workpiece characteristic, etc.],
for proving conformance or non-conformance with the specification.
The iterative method is based basically on an upper bound strategy, i.e. overestimation of the uncertainty at all
levels, but the iterations control the amount of overestimation. Intentional overestimation — and not under-
estimation — is necessary to prevent wrong decisions based on measurement results. The amount of
overestimation shall be controlled by economical evaluation of the situation.
The iterative method is a tool to maximize profit and minimize cost in the metrological activities of a company. The
iterative method/procedure is economically self-adjusting and is also a tool to change/reduce existing uncertainty in
measurement with the aim of reducing cost in metrology (manufacture). The iterative method makes it possible to
compromise between risk, effort and cost in uncertainty estimation and budgeting.
© ISO 1999 – All rights reserved 1

2 Normative references
The following normative documents contain provisions which, through reference in this text, constitute provisions of
this Technical Specification. For dated references, subsequent amendments to, or revisions of, any of these
publications do not apply. However, parties to agreements based on this Technical Specification are encouraged to
investigate the possibility of applying the most recent editions of the normative documents indicated below. For
undated references, the latest edition of the normative document referred to applies. Members of ISO and IEC
maintain registers of currently valid International Standards.
ISO 1:1975, Standard reference temperature for industrial length measurements.
ISO 4288:1996, Geometrical Product Specifications (GPS) — Surface texture: Profile method — Rules and
procedures for the assessment of surface texture.
ISO 9001:1994, Quality systems — Model for quality systems in design, development, production, installation and
servicing.
ISO 9004-1:1994, Quality management and quality system elements — Part 1: Guidelines.
ISO 14253-1:1998, Geometrical Product Specification (GPS) — Inspection by measurement of workpieces and
measuring instruments — Part 1: Decision rules for proving conformance or non-conformance with specifications.
1)
ISO 14253-3:— , Geometrical Product Specification (GPS) — Inspection by measurement of workpieces and
measuring instruments — Part 3: Procedures for evaluating the integrity of uncertainty of measurement values.
ISO 14660-1:1999, Geometrical Product Specification (GPS) — Geometric features — Part 1: General terms and
definitions.
Guide to the expression of uncertainty in measurement (GUM).BIPM, IEC, IFCC,ISO,IUPAC,IUPAP,OIML,
1st edition, 1995.
International Vocabulary of Basic and General Terms in Metrology (VIM). BIPM, IEC, IFCC, ISO, IUPAC, IUPAP,
OIML, 2nd edition, 1993.
3 Terms and definitions
For the purposes of this Technical Specification, the terms and definitions given in ISO 14253-1, ISO 14660-1, VIM,
GUM and the following apply.
3.1
black box model for uncertainty estimation
method of/model for uncertainty estimation in which the output value of a measurement is obtained in the same unit
as the input (stimuli), rather than by measurement of other quantities functionally related to the measurand
NOTE 1 In the black box model — in this Technical Specification — the uncertainty components are assumed additive, the
influence quantities is transformed to the unit of the measurand and the sensitivity coefficients are equal to 1.
NOTE 2 In many cases a complex method of measurement may be looked upon as one simple black box with stimulus in
and result out from the black box. When a black box is opened, it may turn out to contain several "smaller" black boxes and/or
several transparent boxes.
NOTE 3 The method of uncertainty estimation remains a black box method even if it is necessary to make supplementary
measurements to determine the values of influence quantities in order to make corresponding corrections.
1) To be published.
2 © ISO 1999 – All rights reserved

3.2
transparent box model for uncertainty estimation
method of/model for uncertainty estimation in which the value of a measurand is obtained by measurement of other
quantities functionally related to the measurand
3.3
measuring task
quantification of a measurand according to its definition
3.4
basic measurement task (basic measurement)
measurement task(s) which form the basis for evaluation of more complicated characteristics of a workpiece or a
measuring equipment
NOTE Examples of a basic measurement are:
a) one of several individual measurements of the deviation from straightness of a feature of a workpiece;
b) one of the individual measurements of error of indication of a micrometer when measuring the range of error of indication.
3.5
overall measurement task
complicated measuring task, which is evaluated on the basis of several and maybe different basic measurements
NOTE Examples of an overall measuring task are:
a) the measurement of straightness of a feature of a workpiece;
b) the range of error of indication of a micrometer.
3.6
expanded uncertainty (of a measurement)
U
[3.16 of ISO 14253-1:1998 and 2.3.5 of GUM:1995]
NOTE U (capital) always indicates expanded uncertainty of measurement.
3.7
true uncertainty
U
A
uncertainty of measurement that would be obtained by a perfect uncertainty estimation
NOTE 1 True uncertainties are by nature indeterminate.
NOTE 2 See also 8.8.
3.8
conventional true uncertainty — GUM uncertainty
U
c
uncertainty of measurement estimated completely according to the more elaborate procedures of GUM
NOTE 1 The conventional true uncertainty of measurement may differ from an uncertainty of measurement estimated
according to this Technical Specification.
NOTE 2 See also 8.8.
© ISO 1999 – All rights reserved 3

3.9
approximated uncertainty
U
EN
uncertainty of measurement estimated by the simplified, iterative method
NOTE 1 The index N indicates that U is assessed by iteration number N. The designation U may be used without indication
EN E
of the iteration number, when it is without importance to know the number of iterations.
NOTE 2 See also 8.8.
3.10
target uncertainty (for a measurement or calibration)
U
T
uncertainty determined as the optimum for the measuring task
NOTE 1 Target uncertainty is the result of a management decision involving e.g. design, manufacturing, quality assurance,
service, marketing, sales and distribution.
NOTE 2 Target uncertainty is determined (optimized) taking into account the specification [tolerance or maximum
permissible error (MPE)], the process capability, cost, criticality and the requirements of 4.11.1, 4.11.2 of ISO 9001:1994, 13.1
of ISO 9004-1:1994 and ISO 14253-1.
NOTE 3 See also 8.8.
3.11
required uncertainty of measurement
U
R
uncertainty required for a given measurement process and task
NOTE See also 6.2. The required uncertainty may be specified by, for example, a customer.
3.12
uncertainty management
process of deriving an adequate measurement procedure from the measuring task and the target uncertainty by
using uncertainty budgeting techniques
3.13
uncertainty budget (for a measurement or calibration)
statement summarizing the estimation of the uncertainty components that contributes to the uncertainty of a result
of a measurement
NOTE 1 The uncertainty of the result of the measurement is unambiguous only when the measurement procedure (including
the measurement object, measurand, measurement method and conditions) is defined.
NOTE 2 The term "budget" is used for the assignment of numerical values to the uncertainty components, their combination
and expansion, based on the measurement procedure, measurement conditions and assumptions.
3.14
uncertainty contributor
xx
source of uncertainty of measurement for a measuring process
3.15
limit value (variation limit) for an uncertainty contributor
a
xx
absolute value of the extreme value(s) of the uncertainty contributor, xx
4 © ISO 1999 – All rights reserved

3.16
uncertainty component
u
xx
standard uncertainty of the uncertainty contributor, xx
NOTE The iteration method uses the designation u for all uncertainty components. This is not consistent with the present
xx
version of GUM which sometimes uses the designation s for uncertainty components evaluated by A evaluation and the
xx
designation u for uncertainty components evaluated by B evaluation.
xx
3.17
influence quantity of a measurement instrument
characteristic of a measuring instrument that affects the result of a measurement performed by the instrument
3.18
influence quantity of a workpiece
characteristic of a workpiece that affects the result of a measurement performed on that workpiece
© ISO 1999 – All rights reserved 5

4 Symbols
For the purposes of this Technical Specification, the generic symbols given in Table 1 apply.
Table 1 — Generic symbols
Symbol Description
a
limit value for a distribution
a
limit value for an error or uncertainty contributor (in the unit of the result of measurement, of the measurand)
xx
a* limit value for an error or uncertainty contributor (in the unit of the influence quantity)
xx
linear coefficient of thermal expansion

b
coefficient for transformation of a to u
xx xx
C correction (value)
d
resolution of a measurement equipment
E
Young's modulus
ER error (value of a measurement)
G function of several measurement values [G( X , X , . X , .)]
1 2 i
h hysteresis value
k
coverage factor
m number of standard deviations in the half of a confidence interval
MR measurement result (value)
n number of .
N number of iterations
Poisson's number

p number of total uncorrelated uncertainty contributors
r number of total correlated uncertainty contributors
correlation coefficient

TV true value of a measurement
u, u standard uncertainty (standard deviation)
i
s
standard deviation of a sample
x
s
x standard deviation of a mean value of a sample
u combined standard uncertainty
c
u standard deviation of uncertainty contributor xx — uncertainty component
xx
U expanded uncertainty of measurement
U true uncertainty of measurement
A
U conventional true uncertainty of measurement
C
U approximated uncertainty of measurement (number of iteration not stated)
E
U approximated uncertainty of measurement of iteration number N
EN
U required uncertainty
R
U target uncertainty
T
U uncertainty value (not estimated according to GUM or this Technical Specification)
V
X measurement result (uncorrected)
X measurement result (in the transparent box model of uncertainty estimation)
i
Y measurement result (corrected)
6 © ISO 1999 – All rights reserved

5 Concept of the iterative GUM-method for estimation of uncertainty of measurement
Applying the GUM method completely one will find a conventional true uncertainty of measurement, U .
C
The simplified, iterative method/procedure of this Technical Specification is to achieve estimated uncertainties of
measurements, U by overestimating the influencing uncertainty components/contributors (U W U ). The process
E E C
of overestimating provides "worst-case-contributions" at the upper bound from each known or predictable
uncertainty contributor, thus ensuring results of estimations "on the safe side", i.e. not underestimating the
uncertainty of measurement. The simplified, iterative method of this Technical Specification is based on the
following:
� all uncertainty contributors are identified;
� it is decided which of the possible corrections shall be made (see 8.4.6);
� the influence on the uncertainty of the result of measurement from each contributor is evaluated as a standard
uncertainty u , called the uncertainty component;
xx
NOTE As a convention in the iterative method the influence of each contributor must be converted into the unit of the
measurand — using relevant physical equations/formulae and sensibility coefficients.
� an iteration process, PUMA (see clause 6);
� the evaluation of each of the uncertainty components (standard uncertainties) u can take place either by type
xx
A-evaluation or by type B-evaluation;
� type B-evaluation is preferred — if possible — in the first iteration in order to get a rough uncertainty estimate
to establish an overview and to save cost;
� the total effect of all contributors (called the combined standard uncertainty) is calculated by the formula:
22 2 2
uu��u�u� .�u (1)
c1xx2 x3 xn
� the formula (1) is only valid for a black box model of the uncertainty estimation and when the components u
xx
are all uncorrelated (for more details and other formulas see 8.6 and 8.7);
� for simplification the only correlation coefficients between contributors considered are
� =1, –1, 0 (2)
if the uncertainty components are not known to be uncorrelated, full correlation is assumed, either � =1 or� 1.
Correlated components are added arithmetically before put into the formula above (see 8.5 and 8.6);
� the expanded uncertainty U is calculated by the formula:
Uk��u (3)
c
where k =2; k is the coverage factor (see also 8.8);
The simplified, iterative method normally will consist of at least two iterations of estimating the components of
uncertainty.
a) The first very rough, quick and cheap iteration has the purpose of identifying the largest components of
uncertainty (see Figure 1);
b) The following iterations — if any — only deal with making more accurate "upper bound" estimates of the largest
components to lower the estimate of the uncertainty (u and U) to a possible acceptable magnitude.
c
© ISO 1999 – All rights reserved 7

The simplified and iterative method may be used for two purposes:
a) Management of the uncertainty of measurement for a result of a given measurement process (can be used for
the results from a known measuring process or for comparison of two or more of such results) — see 6.2.
b) Uncertainty management for a measuring process. Development of an adequate measuring process i.e.
U u U — see 6.3.
E T
6 Procedure for Uncertainty MAnagement — PUMA
6.1 General
The prerequisite for uncertainty budgeting and management is a clearly identified and defined measuring task; i.e.
the measurand to be quantified (a GPS characteristic of a workpiece or a metrological characteristic of a GPS
measuring equipment). The uncertainty of measurement is a measure of the quality of the measured value
according to the definitions of a GPS characteristic of the workpiece or a metrological characteristic of the GPS
measuring equipment given in GPS standards.
GPS standards define the "conventional true values" (see 1.20 of VIM:1993) of the characteristics to be measured
by chains of standards and global standards (see ISO/TR 14638). GPS standards in many cases also define the
ideal — or conventional true — principle of measurement (see 2.3 of VIM:1993), method of measurement (see 2.4
of VIM:1993), measurement procedure (see 2.5 of VIM:1003) and Standard "reference conditions" (see 5.7 of
VIM:1993).
Deviations from the standardized conventional true values of the characteristics, etc. (the ideal operator) are
contributing to the uncertainty of measurement.
6.2 Uncertainty management for a given measurement process
Management of the uncertainty of measurement for a given measuring task (box 1 of Figure 1) and for an existing
measurement process is illustrated in Figure 1. The principle of measurement (box 3), measurement method
(box 4), measurement procedure (box 5) and measurement conditions (box 6) are fixed and given or decided in this
case, and cannot be changed. The only task is to evaluate the consequence on the uncertainty of measurement. A
required U may be given or decided.
R
Using the iterative GUM method the first iteration is only for orientation, and to look for the dominant uncertainty
contributors. The only thing to do — in the management process in this case — is to refine the estimation of the
dominant contributors to come closer to a true estimate of the uncertainty components thus avoiding a too big
overestimate — if necessary.
Figure 1 — Uncertainty management for a result of measurement from a given measurement process
8 © ISO 1999 – All rights reserved

The procedure is as follows:
a) make a first iteration based preferably on a black box model of the uncertainty estimation process and set up a
preliminary uncertainty budget (boxes 7 to 9) leading to the first rough estimate of the expanded uncertainty,
U (box 10). For details about uncertainty estimation see 9. All estimates of uncertainties U are performed
E1 EN
as upper bound estimates;
b) compare the first estimated uncertainty, U , with the required uncertainty U (box A) for the actual measuring
E1 R
task
1) If U is acceptable (i.e. if U u U ), then the uncertainty budget of the first iteration has proven that the
E1 E1 R
given measurement procedure is adequate for the measuring task (box 11);
2) If U is not acceptable (i.e. if U > U ) or if there is no required uncertainty, but a lower and more true
E1 E1 R
value is desired, the iteration process continues;
c) before the new iteration, analyze the relative magnitude of the uncertainty contributors. In many cases a few
uncertainty components dominate the combined standard uncertainty and expanded uncertainty;
d) change the assumptions or improve the knowledge about the uncertainty components to make a more
accurate (see 3.5 of VIM:1993) upper bound estimation of the largest (dominant) uncertainty components (box
12).
Change to a more detailed model of the uncertainty estimation process or a higher resolution of the measuring
process (box 12);
e) make the second iteration of the uncertainty budget (boxes 7 to 9) leading to the second, lower and more
accurate (see 3.5 of VIM:1993) upper bound estimate of the uncertainty of measurement, U (box 10);
E2
f) compare the second estimated uncertainty U (box A) with uncertainty required U for the actual measuring
E2 R
task
1) if U is acceptable (i.e. if U u U ), then the uncertainty budget of the second iteration has proven that
E2 E2 R
the given measurement procedure is adequate to the measuring task (box 11);
2) if U is not acceptable (i.e. if U > U ), or if there is no required uncertainty, but a lower and more true
E2 E2 R
value is desired, then a third (and possibly more) iteration(s) is (are) needed. Repeat the analysis of the
uncertainty contributors [additional changes of assumptions, improve in knowledge, changes in modelling,
etc. (box 12)] and concentrate on the currently largest uncertainty contributors;
g) when all possibilities have been used for making more accurate (lower) upper bound estimates of the
measuring uncertainties without coming to an acceptable measuring uncertainty U u U , then it is proven,
EN R
that it is not possible to fulfil the given requirement U .
R
6.3 Uncertainty management for design and development of a measurement process/procedure
Uncertainty management in this case is performed to develop an adequate measurement procedure [measurement
of the geometrical characteristics of a workpiece or the metrological characteristics of a measuring equipment
(calibration)]. Uncertainty management is performed on the basis of a defined measuring task (box 1 in Figure 2)
and a given target uncertainty, U (box 2 in Figure 2). Definition of the measuring task and target uncertainty are
T
company policy decisions to be made at a sufficiently high management level. An adequate measurement
procedure is a procedure which results in an estimated uncertainty of measurement less than or equal to the target
uncertainty. If the estimated uncertainty of measurement is much less than the target uncertainty, the measurement
procedure may not be (economically) optimal for performing the measuring task (i.e. the measurement process is
too costly).
© ISO 1999 – All rights reserved 9

The PUMA, based on a given measuring task (box 1) and a given target uncertainty U (box 2), includes the
T
following (see Figure 2):
a) choose the principle of measurement (box 3) on the basis of experience and possible measurement
instruments present in the company;
b) set up and document a preliminary method of measurement (box 4), measurement procedure (box 5) and
measurement conditions (box 6) on the basis of experience and known possibilities in the company;
c) make a first iteration based preferably on a black box model of the uncertainty estimation process and set up a
preliminary uncertainty budget (boxes 7 to 9) leading to the first rough estimate of the expanded uncertainty,
U (box 10). For details about uncertainty estimation see clause 9. All estimates of uncertainties U are
E1 EN
performed as upper bound estimates;
d) compare the first estimated uncertainty, U , with the given target uncertainty, U (box A);
E1 T
1) if U is acceptable (i.e. if U u U ), then the uncertainty budget of the first iteration has proven that the
E1 E1 T
measurement procedure is adequate for the measuring task (box 11);
2) if U << U , then the measurement procedure is technically acceptable, but a possibility may exist to
E1 T
change the method and/or the procedure (box 13) in order to make the measuring process more cost
effective while increasing the uncertainty. A new iteration is then needed to estimate the resulting
measurement uncertainty, U (box 10);
E2
3) if U is not acceptable (i.e. if U > U ), the iteration process continues, or it is concluded that no
E1 E1 T
adequate measurement procedure is possible;
e) before the new iteration, analyze the relative magnitude of the uncertainty contributors. In many cases a few
uncertainty components pre-dominate the combined standard uncertainty and expanded uncertainty;
f) if U > U , then change the assumptions, the modelling or increase the knowledge about the uncertainty
E1 T
components (box 12) to make a more accurate (see 3.5 of VIM:1993) upper bound estimation of the largest
(dominant) uncertainty components;
g) make the second iteration of the uncertainty budget (boxes 7 to 9) leading to the second, lower and more
accurate (see 3.5 of VIM:1993) upper bound estimate of the uncertainty of measurement, U (box 10);
E2
h) compare the second estimated uncertainty U with the given target uncertainty, U (box A);
E2 T
1) if U is acceptable (i.e. if U u U ), then the uncertainty budget of the second iteration has proven that
E2 E2 T
the measurement procedure is adequate for the measuring task (box 11);
2) if U is not acceptable (i.e. if U > U ) then a third (and possibly more) iteration(s) is (are) needed.
E2 E2 T
Repeat the analysis of the uncertainty contributors (additional changes of assumptions, modelling and
increase in knowledge (box 12)) and concentrate on the currently largest uncertainty contributors;
i) when all possibilities has been used for making more accurate (lower) upper bound estimates of the measuring
uncertainties without coming to an acceptable measuring uncertainty U u U , then a change of the
EN T
measurement method or the measurement procedure or the conditions of measurement (box 13) is needed to
(possibly) bring down the magnitude of the estimated uncertainty, U . The iteration procedure starts again
EN
with a first iteration;
j) if changes in the measurement method or the measurement procedure or conditions (box 13) do not lead to an
acceptable uncertainty of measurement, the final possibility is to change the principle of measurement (box 14)
and start the above mentioned procedure again;
10 © ISO 1999 – All rights reserved

k) if change of the measuring principle and the related iterations described above do not lead to an acceptable
uncertainty of measurement the ultimate possibility is to change the measuring task and/or target uncertainty
(box 15) and start the above mentioned procedure again;
l) if change of measuring task or target uncertainty is not possible, it is demonstrated, that no adequate
measurement procedure exists (box 16).
© ISO 1999 – All rights reserved 11

Figure 2 — Procedure for Uncertainty of Measurement MAnagement (PUMA) for a measurement
process/procedure
12 © ISO 1999 – All rights reserved

7 Sources of errors and uncertainty of measurement
7.1 Types of errors
Different types of errors regularly shows up in measurement results.
� systematic errors;
� random errors;
� drift;
� outliers.
All errors are by nature systematic. When we see errors as non-systematic it is because the reason for the error is
not looked for or because the level of resolution is not sufficient. Systematic errors may be characterised by size
and sign (+ or�).
ER = MR� TV
where
ER is the error,
MR is the measurement result;
TV is thetruevalue.
Random errors are systematic errors caused by non-controlled random influence quantities. Random errors may be
characterized by the standard deviation and the type of distribution. The mean value of the random errors is often
considered as a basis for the evaluation of the systematic error (see Figure 3).
Key
1 Outlier
2 Dispersion 1
3 Dispersion 2
4 Systematic error 1
5 Systematic error 2
6 True value
Figure 3 — Types of errors in results of measurements
© ISO 1999 – All rights reserved 13

Drift is caused by a systematic influence of non-controlled influence quantities. Drift is often a time effect or a wear
effect. Drift may be characterized by change per unit time or per amount of use.
Outliers are caused by not repeatable incidents in the measurement. Noise — electrical or mechanical — may
result in outliers. A frequent reason for outliers is human mistakes as reading and writing errors or wrong handling
of measuring equipment. Outliers are impossible to characterize in advance.
Errors or uncertainties in a measuring process will be a mix of known and unknown errors from a number of
sources or error contributors.
The sources or contributors are not the same in each case, and the sum of the components are not the same.
It is still possible to make a systematic approach. There are always several sources or a combined effect of the ten
different ones indicated in Figure 4.
In the following, examples and further details about each of the ten contributors are given.
What is often difficult is that each of the contributors may act individually on the result of measurement. But in many
cases they even interfere with each other and cause additional errors and uncertainty.
Figure 4 and the following non-exhaustive lists (see 7.2 to 7.11) shall be used for getting ideas in a systematic way
when making uncertainty budgets. In each case the evaluation of the actual error/uncertainty component needs
knowledge about physics and/or experience in metrology.
In uncertainty budgets the uncertainty contributors and the uncertainty components may be grouped for
convenience.
Figure 4 —Uncertainty contributors in measurement
14 © ISO 1999 – All rights reserved

7.2 Environment for the measurement
In most cases — especially in GPS measurements — the temperature is the main uncertainty contributor of the
environment. Other uncertainty contributors may be:
� Temperature: absolute temperature, time � Gravity
variance, spatial gradient
� Electromagnetic interference
� Vibration/noise
� Transients in the power supply
� Humidity
� Pressured air (e.g. air bearings)
� Contamination
� Heat radiation
� Illumination
� Workpiece
� Ambient pressure
� Scale
� Air composition
� Instrument thermal equilibrium
� Air flow
7.3 Reference element of measurement equipment
The measuring equipment is divided into "reference element" and the "rest of the equipment", and it often pays to
look at the equipment that way.
� Stability � CCD-techniques
� Scale mark quality � Uncertainty of the calibration
� Temperature expansion coefficient � Resolution of the main scale (analogue or digital)
� Physical principle: line scale, optical digital scale, � Time since last calibration
magnetic digital scale, spindle, rack & pinion,
� Wavelength error
interferometer
7.4 Measurement equipment
� Interpretation system � Reading system
� Magnification, electrical or mechanical � Linear coefficient for thermal expansion
� Error wavelength � Temperature stability/sensitivity
� Zero-point stability � Parallaxes
� Force stability/absolute force � Time since last calibration
� Hysteresis � Response characteristic
� Guides/slideways � Interpolation system, error wavelength
� Probe system � Interpolation resolution
� Geometrical imperfections � Digitization
� Stiffness/rigidity
© ISO 1999 – All rights reserved 15

7.5 Measurement setup (excluding the placement and clamping of the workpiece)
In many cases there is no setup; the measurement equipment can measure "alone".
� Cosine errors and sine errors � Form deviation of tip
� Abbe principle � Stiffness of the probe system
� Temperature sensitivity � Optical aperture
� Stiffness/rigidity � Interaction between workpiece and setup
� Tip radius � Warming up
7.6 Software and calculations
Observe that even the number of digits or decimals can have an influence!
� Rounding/Quantification � Filtering
� Algorithms � Correction of algorithm/Certification of algorithm
� Implementation of algorithms � Interpolation/extrapolation
� Number of significant digits in the computation � Outlier handling
� Sampling
7.7 Metrologist
The human being is not stable; there is a difference from day to day and often a rather large change during the
day.
� Education � Knowledge (precision, appreciation)
� Experience � Honesty
� Training � Dedication
� Physical disadvantages/ability
7.8 Measurement object, workpiece or measuring instrument characteristic
� Surface roughness � Magnetism
� Form deviations � Hygroscopic characteristic of the material
� E-modulus (Young's modulus) � Ageing
� Stiffness beyond E-modulus � Cleanliness
� Temperature expansion coefficient � Temperature
� Conductivity � Internal stress
� Weight � Creep characteristics
� Size � Workpiece distortion due to clamping
� Shape � Orientation
16 © ISO 1999 – All rights reserved

7.9 Definition of the GPS characteristic, workpiece or measuring instrument characteristic
� Datum � ISO 4288
� Reference system � Chain link 3 and 4 deviations (ISO/TR 14638)
� Degrees of freedom � Distance
� Toleranced feature � Angle
7.10 Measuring procedure
� Conditioning � Number of operators
� Number of measurements � Strategy
� Order of measurements � Clamping
� Duration of measurements � Fixturing
� Choice of principle of measurement � Number of points
� Alignment � Probing principle and strategy
� Choice of reference — reference item (standard) � Alignment of probing system
and value — relative to the measured value
� Drift check
� Choice of apparatus
� Reversal measurements
� Choice of metrologist
� Multiple redundancy, error separation
7.11 Physical constants and conversion factors
� Knowledge of the correct physical values of, for example, material properties (workpiece, measuring
instrument, ambient air, etc.)
8 Tools for the estimation of uncertainty components, standard uncertainty and
expanded uncertainty
8.1 Estimation of uncertainty components
Estimation of uncertainty components can be done in two different ways. Type A evaluation and type B evaluation.
Type A-evaluation is evaluation of uncertainty components, u , using statistical means. Type B evaluation is
xx
evaluation of uncertainty components, u , by any other means than statistical.
xx
Type A-evaluation will in most cases result in more accurate estimates of uncertainty components than type B-
evaluation. In many cases Type B evaluation will result in sufficiently accurate estimations of uncertainty
components.
Therefore, Type B evaluation shall be chosen in the itera
...


SLOVENSKI STANDARD
01-julij-2002
6SHFLILNDFLMDJHRPHWULMVNLKYHOLþLQL]GHOND3UHLVNDYH]PHUMHQMHPL]GHONDLQ
PHULOQDRSUHPDGHO1DYRGLOD]DRFHQMHYDQMHQHJRWRYRVWLSULPHULWYDK
JHRPHWULMVNLKYHOLþLQL]GHONRYSULXPHUMDQMXPHULOQHRSUHPHLQSULSRWUMHYDQMX
SURL]YRGRY
Geometrical Product Specifications (GPS) -- Inspection by measurement of workpieces
and measuring equipment -- Part 2: Guide to the estimation of uncertainty in GPS
measurement, in calibration of measuring equipment and in product verification
Spécification géométrique des produits (GPS) -- Vérification par la mesure des pièces et
des équipements de mesure -- Partie 2: Guide pour l'estimation de l'incertitude dans les
mesures GPS, dans l'étalonnage des équipements de mesure et dans la vérification des
produits
Ta slovenski standard je istoveten z: ISO/TS 14253-2:1999
ICS:
17.040.20 Lastnosti površin Properties of surfaces
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

TECHNICAL ISO/TS
SPECIFICATION 14253-2
First edition
1999-12-01
Geometrical Product Specifications
(GPS) — Inspection by measurement of
workpieces and measuring equipment —
Part 2:
Guide to the estimation of uncertainty
in GPS measurement, in calibration
of measuring equipment and in product
verification
Spécification géométrique des produits (GPS) — Vérification par la mesure
des pièces et des équipements de mesure —
Partie 2: Guide pour l'estimation de l'incertitude dans les mesures GPS,
dans l'étalonnage des équipements de mesure et dans la vérification
des produits
Reference number
©
ISO 1999
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but shall not
be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In downloading this
file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat accepts no liability in this
area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation parameters
were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In the unlikely event
that a problem relating to it is found, please inform the Central Secretariat at the address given below.
© ISO 1999
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means, electronic
or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or ISO's member body
in the country of the requester.
ISO copyright office
Case postale 56 � CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 734 10 79
E-mail copyright@iso.ch
Web www.iso.ch
Printed in Switzerland
ii © ISO 1999 – All rights reserved

Contents Page
Foreword.iv
Introduction.v
1 Scope .1
2 Normative references .2
3 Terms and definitions .2
4 Symbols .6
5 Concept of the iterative GUM-method for estimation of uncertainty of measurement .7
6 Procedure for Uncertainty MAnagement — PUMA .8
7 Sources of errors and uncertainty of measurement.13
8 Tools for the estimation of uncertainty components, standard uncertainty and expanded
uncertainty.17
9 Practical estimation of uncertainty — Uncertainty budgeting with PUMA.26
10 Applications .30
Annex A (informative) Example of uncertainty budgets — Calibration of a setting ring.34
Annex B (informative) Example of uncertainty budgets — Design of a calibration hierarchy.41
Annex C (informative) Example of uncertainty budgets — Measurement of roundness .65
Annex D (informative) Relation to the GPS matrix model.71
Bibliography.73
© ISO 1999 – All rights reserved iii

Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies (ISO
member bodies). The work of preparing International Standards is normally carried out through ISO technical
committees. Each member body interested in a subject for which a technical committee has been established has
the right to be represented on that committee. International organizations, governmental and non-governmental, in
liaison with ISO, also take part in the work. ISO collaborates closely with the International Electrotechnical
Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 3.
The main task of technical committees is to prepare International Standards. Draft International Standards adopted
by the technical committees are circulated to the member bodies for voting. Publication as an International
Standard requires approval by at least 75 % of the member bodies casting a vote.
In other circumstances, particularly when there is an urgent market requirement for such documents, a technical
committee may decide to publish other types of normative document:
— an ISO Publicly Available Specification (ISO/PAS) represents an agreement between technical experts in an
ISO working group and is accepted for publication if it is approved by more than 50 % of the members of the
parent committee casting a vote;
— an ISO Technical Specification (ISO/TS) represents an agreement between the members of a technical
committee and is accepted for publication if it is approved by 2/3 of the members of the committee casting a
vote.
An ISO/PAS or ISO/TS is reviewed every three years with a view to deciding whether it can be transformed into an
International Standard.
Attention is drawn to the possibility that some of the elements of this Technical Specification may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO/TS 14253-2 was prepared by Technical Committee ISO/TC 213, Dimensional and geometrical product
specifications and verification.
ISO 14253 consists of the following parts, under the general title Geometrical product specifications (GPS) —
Inspection by measurement of workpieces and measuring equipment:
� Part 1: Decision rules for proving conformance or non-conformance with specification
� Part 2: Guide to the estimation of uncertainty in GPS measurement, in calibration of measuring equipment and
in product verification [Technical Specification]
� Part 3: Procedures for evaluating the integrity of uncertainty in measurement values
Annexes A to D of this Technical Specification are for information only.
iv © ISO 1999 – All rights reserved

Introduction
This Technical Specification is a global GPS technical report (see ISO/TR 14638:1995). This global GPS Technical
Report influences chain link 4, 5 and 6 in all chains of standards.
For more detailed information of the relation of this report to other standards and the GPS matrix model, see
annex D.
This Technical Specification is developed to support ISO 14253-1. This Technical Specification establishes a
simplified, iterative procedure of the concept and the way to evaluate and determine uncertainty (standard
uncertainty and expanded uncertainty) of measurement, and the recommendations of the format to document and
report the uncertainty of measurement information as given in "Guide to the expression of uncertainty in
measurement" (GUM). In most cases only very limited resources are necessary to estimate uncertainty of
measurement by this simplified, iterative procedure, but the procedure may lead to a slight overestimation of the
uncertainty of measurement. If a more accurate estimation of the uncertainty of measurement is needed, the more
elaborated procedures of the GUM must be applied.
This simplified, iterative procedure of the GUM methods is intended for GPS measurements, but may be used in
other areas of industrial (applied) metrology.
Uncertainty of measurement and the concept of handling uncertainty of measurement being of importance to all the
technical functions in a company, this Technical Specification relates to e.g. management function, design and
development function, manufacture function, quality assurance function, metrology function, etc.
This Technical Specification is of special importance in relation to ISO 9000 quality assurance systems, where
it is a requirement that the uncertainty of measurement is known [e.g. 4.11.1, 4.11.2 a) and 4.11.2 b) of
ISO 9001:1994].
In this Technical Specification the uncertainty of the result of a process of calibration and a process of
measurement is handled in the same way:
� calibration is treated as "measurement of metrological characteristics of a measuring equipment or a
measurement standard";
� measurement is treated as "measurement of geometrical characteristics of a workpiece".
Therefore, in most cases no distinction is made in the text between measurement and calibration. The term
"measurement" is used as a synonym for both.
© ISO 1999 – All rights reserved v

TECHNICAL SPECIFICATION ISO/TS 14253-2:1999(E)
Geometrical product specifications (GPS) — Inspection by
measurement of workpieces and measuring equipment —
Part 2:
Guide to the estimation of uncertainty in GPS measurement, in
calibration of measuring equipment and in product verification
1 Scope
This Technical Specification gives guidance on the implementation of the concept of "Guide to the estimation of
uncertainty in measurement" (in short GUM) to be applied in industry for the calibration of (measurement)
standards and measuring equipment in the field of GPS and the measurement of workpiece GPS-characteristics.
The aim is to promote full information on how to achieve uncertainty statements and provide the basis for
international comparison of results of measurements and their uncertainties (relationship between purchaser and
supplier).
This Technical Specification is intended to support ISO 14253-1. This Technical Specification and ISO 14253-1 are
beneficial to all technical functions in a company in the interpretation of GPS specifications (i.e. tolerances of
workpiece characteristics and values of maximum permissible errors (MPE) for metrological characteristics of
measuring equipment).
This Technical Specification introduces the Procedure for Uncertainty MAnagement (PUMA), which is a practical,
iterative procedure based on the GUM for estimating uncertainty of measurement without changing the basic
concepts of the GUM and is intended to be used generally for estimating uncertainty of measurement and giving
statements of uncertainty for:
� single results of measurement;
� comparison of two or more results of measurement;
� comparison of results of measurement — from one or more workpieces or pieces of measurement equipment
— with given specifications [i.e. maximum permissible errors (MPE) for a metrological characteristic of a
measurement instrument or measurement standard, and tolerance limits for a workpiece characteristic, etc.],
for proving conformance or non-conformance with the specification.
The iterative method is based basically on an upper bound strategy, i.e. overestimation of the uncertainty at all
levels, but the iterations control the amount of overestimation. Intentional overestimation — and not under-
estimation — is necessary to prevent wrong decisions based on measurement results. The amount of
overestimation shall be controlled by economical evaluation of the situation.
The iterative method is a tool to maximize profit and minimize cost in the metrological activities of a company. The
iterative method/procedure is economically self-adjusting and is also a tool to change/reduce existing uncertainty in
measurement with the aim of reducing cost in metrology (manufacture). The iterative method makes it possible to
compromise between risk, effort and cost in uncertainty estimation and budgeting.
© ISO 1999 – All rights reserved 1

2 Normative references
The following normative documents contain provisions which, through reference in this text, constitute provisions of
this Technical Specification. For dated references, subsequent amendments to, or revisions of, any of these
publications do not apply. However, parties to agreements based on this Technical Specification are encouraged to
investigate the possibility of applying the most recent editions of the normative documents indicated below. For
undated references, the latest edition of the normative document referred to applies. Members of ISO and IEC
maintain registers of currently valid International Standards.
ISO 1:1975, Standard reference temperature for industrial length measurements.
ISO 4288:1996, Geometrical Product Specifications (GPS) — Surface texture: Profile method — Rules and
procedures for the assessment of surface texture.
ISO 9001:1994, Quality systems — Model for quality systems in design, development, production, installation and
servicing.
ISO 9004-1:1994, Quality management and quality system elements — Part 1: Guidelines.
ISO 14253-1:1998, Geometrical Product Specification (GPS) — Inspection by measurement of workpieces and
measuring instruments — Part 1: Decision rules for proving conformance or non-conformance with specifications.
1)
ISO 14253-3:— , Geometrical Product Specification (GPS) — Inspection by measurement of workpieces and
measuring instruments — Part 3: Procedures for evaluating the integrity of uncertainty of measurement values.
ISO 14660-1:1999, Geometrical Product Specification (GPS) — Geometric features — Part 1: General terms and
definitions.
Guide to the expression of uncertainty in measurement (GUM).BIPM, IEC, IFCC,ISO,IUPAC,IUPAP,OIML,
1st edition, 1995.
International Vocabulary of Basic and General Terms in Metrology (VIM). BIPM, IEC, IFCC, ISO, IUPAC, IUPAP,
OIML, 2nd edition, 1993.
3 Terms and definitions
For the purposes of this Technical Specification, the terms and definitions given in ISO 14253-1, ISO 14660-1, VIM,
GUM and the following apply.
3.1
black box model for uncertainty estimation
method of/model for uncertainty estimation in which the output value of a measurement is obtained in the same unit
as the input (stimuli), rather than by measurement of other quantities functionally related to the measurand
NOTE 1 In the black box model — in this Technical Specification — the uncertainty components are assumed additive, the
influence quantities is transformed to the unit of the measurand and the sensitivity coefficients are equal to 1.
NOTE 2 In many cases a complex method of measurement may be looked upon as one simple black box with stimulus in
and result out from the black box. When a black box is opened, it may turn out to contain several "smaller" black boxes and/or
several transparent boxes.
NOTE 3 The method of uncertainty estimation remains a black box method even if it is necessary to make supplementary
measurements to determine the values of influence quantities in order to make corresponding corrections.
1) To be published.
2 © ISO 1999 – All rights reserved

3.2
transparent box model for uncertainty estimation
method of/model for uncertainty estimation in which the value of a measurand is obtained by measurement of other
quantities functionally related to the measurand
3.3
measuring task
quantification of a measurand according to its definition
3.4
basic measurement task (basic measurement)
measurement task(s) which form the basis for evaluation of more complicated characteristics of a workpiece or a
measuring equipment
NOTE Examples of a basic measurement are:
a) one of several individual measurements of the deviation from straightness of a feature of a workpiece;
b) one of the individual measurements of error of indication of a micrometer when measuring the range of error of indication.
3.5
overall measurement task
complicated measuring task, which is evaluated on the basis of several and maybe different basic measurements
NOTE Examples of an overall measuring task are:
a) the measurement of straightness of a feature of a workpiece;
b) the range of error of indication of a micrometer.
3.6
expanded uncertainty (of a measurement)
U
[3.16 of ISO 14253-1:1998 and 2.3.5 of GUM:1995]
NOTE U (capital) always indicates expanded uncertainty of measurement.
3.7
true uncertainty
U
A
uncertainty of measurement that would be obtained by a perfect uncertainty estimation
NOTE 1 True uncertainties are by nature indeterminate.
NOTE 2 See also 8.8.
3.8
conventional true uncertainty — GUM uncertainty
U
c
uncertainty of measurement estimated completely according to the more elaborate procedures of GUM
NOTE 1 The conventional true uncertainty of measurement may differ from an uncertainty of measurement estimated
according to this Technical Specification.
NOTE 2 See also 8.8.
© ISO 1999 – All rights reserved 3

3.9
approximated uncertainty
U
EN
uncertainty of measurement estimated by the simplified, iterative method
NOTE 1 The index N indicates that U is assessed by iteration number N. The designation U may be used without indication
EN E
of the iteration number, when it is without importance to know the number of iterations.
NOTE 2 See also 8.8.
3.10
target uncertainty (for a measurement or calibration)
U
T
uncertainty determined as the optimum for the measuring task
NOTE 1 Target uncertainty is the result of a management decision involving e.g. design, manufacturing, quality assurance,
service, marketing, sales and distribution.
NOTE 2 Target uncertainty is determined (optimized) taking into account the specification [tolerance or maximum
permissible error (MPE)], the process capability, cost, criticality and the requirements of 4.11.1, 4.11.2 of ISO 9001:1994, 13.1
of ISO 9004-1:1994 and ISO 14253-1.
NOTE 3 See also 8.8.
3.11
required uncertainty of measurement
U
R
uncertainty required for a given measurement process and task
NOTE See also 6.2. The required uncertainty may be specified by, for example, a customer.
3.12
uncertainty management
process of deriving an adequate measurement procedure from the measuring task and the target uncertainty by
using uncertainty budgeting techniques
3.13
uncertainty budget (for a measurement or calibration)
statement summarizing the estimation of the uncertainty components that contributes to the uncertainty of a result
of a measurement
NOTE 1 The uncertainty of the result of the measurement is unambiguous only when the measurement procedure (including
the measurement object, measurand, measurement method and conditions) is defined.
NOTE 2 The term "budget" is used for the assignment of numerical values to the uncertainty components, their combination
and expansion, based on the measurement procedure, measurement conditions and assumptions.
3.14
uncertainty contributor
xx
source of uncertainty of measurement for a measuring process
3.15
limit value (variation limit) for an uncertainty contributor
a
xx
absolute value of the extreme value(s) of the uncertainty contributor, xx
4 © ISO 1999 – All rights reserved

3.16
uncertainty component
u
xx
standard uncertainty of the uncertainty contributor, xx
NOTE The iteration method uses the designation u for all uncertainty components. This is not consistent with the present
xx
version of GUM which sometimes uses the designation s for uncertainty components evaluated by A evaluation and the
xx
designation u for uncertainty components evaluated by B evaluation.
xx
3.17
influence quantity of a measurement instrument
characteristic of a measuring instrument that affects the result of a measurement performed by the instrument
3.18
influence quantity of a workpiece
characteristic of a workpiece that affects the result of a measurement performed on that workpiece
© ISO 1999 – All rights reserved 5

4 Symbols
For the purposes of this Technical Specification, the generic symbols given in Table 1 apply.
Table 1 — Generic symbols
Symbol Description
a
limit value for a distribution
a
limit value for an error or uncertainty contributor (in the unit of the result of measurement, of the measurand)
xx
a* limit value for an error or uncertainty contributor (in the unit of the influence quantity)
xx
linear coefficient of thermal expansion

b
coefficient for transformation of a to u
xx xx
C correction (value)
d
resolution of a measurement equipment
E
Young's modulus
ER error (value of a measurement)
G function of several measurement values [G( X , X , . X , .)]
1 2 i
h hysteresis value
k
coverage factor
m number of standard deviations in the half of a confidence interval
MR measurement result (value)
n number of .
N number of iterations
Poisson's number

p number of total uncorrelated uncertainty contributors
r number of total correlated uncertainty contributors
correlation coefficient

TV true value of a measurement
u, u standard uncertainty (standard deviation)
i
s
standard deviation of a sample
x
s
x standard deviation of a mean value of a sample
u combined standard uncertainty
c
u standard deviation of uncertainty contributor xx — uncertainty component
xx
U expanded uncertainty of measurement
U true uncertainty of measurement
A
U conventional true uncertainty of measurement
C
U approximated uncertainty of measurement (number of iteration not stated)
E
U approximated uncertainty of measurement of iteration number N
EN
U required uncertainty
R
U target uncertainty
T
U uncertainty value (not estimated according to GUM or this Technical Specification)
V
X measurement result (uncorrected)
X measurement result (in the transparent box model of uncertainty estimation)
i
Y measurement result (corrected)
6 © ISO 1999 – All rights reserved

5 Concept of the iterative GUM-method for estimation of uncertainty of measurement
Applying the GUM method completely one will find a conventional true uncertainty of measurement, U .
C
The simplified, iterative method/procedure of this Technical Specification is to achieve estimated uncertainties of
measurements, U by overestimating the influencing uncertainty components/contributors (U W U ). The process
E E C
of overestimating provides "worst-case-contributions" at the upper bound from each known or predictable
uncertainty contributor, thus ensuring results of estimations "on the safe side", i.e. not underestimating the
uncertainty of measurement. The simplified, iterative method of this Technical Specification is based on the
following:
� all uncertainty contributors are identified;
� it is decided which of the possible corrections shall be made (see 8.4.6);
� the influence on the uncertainty of the result of measurement from each contributor is evaluated as a standard
uncertainty u , called the uncertainty component;
xx
NOTE As a convention in the iterative method the influence of each contributor must be converted into the unit of the
measurand — using relevant physical equations/formulae and sensibility coefficients.
� an iteration process, PUMA (see clause 6);
� the evaluation of each of the uncertainty components (standard uncertainties) u can take place either by type
xx
A-evaluation or by type B-evaluation;
� type B-evaluation is preferred — if possible — in the first iteration in order to get a rough uncertainty estimate
to establish an overview and to save cost;
� the total effect of all contributors (called the combined standard uncertainty) is calculated by the formula:
22 2 2
uu��u�u� .�u (1)
c1xx2 x3 xn
� the formula (1) is only valid for a black box model of the uncertainty estimation and when the components u
xx
are all uncorrelated (for more details and other formulas see 8.6 and 8.7);
� for simplification the only correlation coefficients between contributors considered are
� =1, –1, 0 (2)
if the uncertainty components are not known to be uncorrelated, full correlation is assumed, either � =1 or� 1.
Correlated components are added arithmetically before put into the formula above (see 8.5 and 8.6);
� the expanded uncertainty U is calculated by the formula:
Uk��u (3)
c
where k =2; k is the coverage factor (see also 8.8);
The simplified, iterative method normally will consist of at least two iterations of estimating the components of
uncertainty.
a) The first very rough, quick and cheap iteration has the purpose of identifying the largest components of
uncertainty (see Figure 1);
b) The following iterations — if any — only deal with making more accurate "upper bound" estimates of the largest
components to lower the estimate of the uncertainty (u and U) to a possible acceptable magnitude.
c
© ISO 1999 – All rights reserved 7

The simplified and iterative method may be used for two purposes:
a) Management of the uncertainty of measurement for a result of a given measurement process (can be used for
the results from a known measuring process or for comparison of two or more of such results) — see 6.2.
b) Uncertainty management for a measuring process. Development of an adequate measuring process i.e.
U u U — see 6.3.
E T
6 Procedure for Uncertainty MAnagement — PUMA
6.1 General
The prerequisite for uncertainty budgeting and management is a clearly identified and defined measuring task; i.e.
the measurand to be quantified (a GPS characteristic of a workpiece or a metrological characteristic of a GPS
measuring equipment). The uncertainty of measurement is a measure of the quality of the measured value
according to the definitions of a GPS characteristic of the workpiece or a metrological characteristic of the GPS
measuring equipment given in GPS standards.
GPS standards define the "conventional true values" (see 1.20 of VIM:1993) of the characteristics to be measured
by chains of standards and global standards (see ISO/TR 14638). GPS standards in many cases also define the
ideal — or conventional true — principle of measurement (see 2.3 of VIM:1993), method of measurement (see 2.4
of VIM:1993), measurement procedure (see 2.5 of VIM:1003) and Standard "reference conditions" (see 5.7 of
VIM:1993).
Deviations from the standardized conventional true values of the characteristics, etc. (the ideal operator) are
contributing to the uncertainty of measurement.
6.2 Uncertainty management for a given measurement process
Management of the uncertainty of measurement for a given measuring task (box 1 of Figure 1) and for an existing
measurement process is illustrated in Figure 1. The principle of measurement (box 3), measurement method
(box 4), measurement procedure (box 5) and measurement conditions (box 6) are fixed and given or decided in this
case, and cannot be changed. The only task is to evaluate the consequence on the uncertainty of measurement. A
required U may be given or decided.
R
Using the iterative GUM method the first iteration is only for orientation, and to look for the dominant uncertainty
contributors. The only thing to do — in the management process in this case — is to refine the estimation of the
dominant contributors to come closer to a true estimate of the uncertainty components thus avoiding a too big
overestimate — if necessary.
Figure 1 — Uncertainty management for a result of measurement from a given measurement process
8 © ISO 1999 – All rights reserved

The procedure is as follows:
a) make a first iteration based preferably on a black box model of the uncertainty estimation process and set up a
preliminary uncertainty budget (boxes 7 to 9) leading to the first rough estimate of the expanded uncertainty,
U (box 10). For details about uncertainty estimation see 9. All estimates of uncertainties U are performed
E1 EN
as upper bound estimates;
b) compare the first estimated uncertainty, U , with the required uncertainty U (box A) for the actual measuring
E1 R
task
1) If U is acceptable (i.e. if U u U ), then the uncertainty budget of the first iteration has proven that the
E1 E1 R
given measurement procedure is adequate for the measuring task (box 11);
2) If U is not acceptable (i.e. if U > U ) or if there is no required uncertainty, but a lower and more true
E1 E1 R
value is desired, the iteration process continues;
c) before the new iteration, analyze the relative magnitude of the uncertainty contributors. In many cases a few
uncertainty components dominate the combined standard uncertainty and expanded uncertainty;
d) change the assumptions or improve the knowledge about the uncertainty components to make a more
accurate (see 3.5 of VIM:1993) upper bound estimation of the largest (dominant) uncertainty components (box
12).
Change to a more detailed model of the uncertainty estimation process or a higher resolution of the measuring
process (box 12);
e) make the second iteration of the uncertainty budget (boxes 7 to 9) leading to the second, lower and more
accurate (see 3.5 of VIM:1993) upper bound estimate of the uncertainty of measurement, U (box 10);
E2
f) compare the second estimated uncertainty U (box A) with uncertainty required U for the actual measuring
E2 R
task
1) if U is acceptable (i.e. if U u U ), then the uncertainty budget of the second iteration has proven that
E2 E2 R
the given measurement procedure is adequate to the measuring task (box 11);
2) if U is not acceptable (i.e. if U > U ), or if there is no required uncertainty, but a lower and more true
E2 E2 R
value is desired, then a third (and possibly more) iteration(s) is (are) needed. Repeat the analysis of the
uncertainty contributors [additional changes of assumptions, improve in knowledge, changes in modelling,
etc. (box 12)] and concentrate on the currently largest uncertainty contributors;
g) when all possibilities have been used for making more accurate (lower) upper bound estimates of the
measuring uncertainties without coming to an acceptable measuring uncertainty U u U , then it is proven,
EN R
that it is not possible to fulfil the given requirement U .
R
6.3 Uncertainty management for design and development of a measurement process/procedure
Uncertainty management in this case is performed to develop an adequate measurement procedure [measurement
of the geometrical characteristics of a workpiece or the metrological characteristics of a measuring equipment
(calibration)]. Uncertainty management is performed on the basis of a defined measuring task (box 1 in Figure 2)
and a given target uncertainty, U (box 2 in Figure 2). Definition of the measuring task and target uncertainty are
T
company policy decisions to be made at a sufficiently high management level. An adequate measurement
procedure is a procedure which results in an estimated uncertainty of measurement less than or equal to the target
uncertainty. If the estimated uncertainty of measurement is much less than the target uncertainty, the measurement
procedure may not be (economically) optimal for performing the measuring task (i.e. the measurement process is
too costly).
© ISO 1999 – All rights reserved 9

The PUMA, based on a given measuring task (box 1) and a given target uncertainty U (box 2), includes the
T
following (see Figure 2):
a) choose the principle of measurement (box 3) on the basis of experience and possible measurement
instruments present in the company;
b) set up and document a preliminary method of measurement (box 4), measurement procedure (box 5) and
measurement conditions (box 6) on the basis of experience and known possibilities in the company;
c) make a first iteration based preferably on a black box model of the uncertainty estimation process and set up a
preliminary uncertainty budget (boxes 7 to 9) leading to the first rough estimate of the expanded uncertainty,
U (box 10). For details about uncertainty estimation see clause 9. All estimates of uncertainties U are
E1 EN
performed as upper bound estimates;
d) compare the first estimated uncertainty, U , with the given target uncertainty, U (box A);
E1 T
1) if U is acceptable (i.e. if U u U ), then the uncertainty budget of the first iteration has proven that the
E1 E1 T
measurement procedure is adequate for the measuring task (box 11);
2) if U << U , then the measurement procedure is technically acceptable, but a possibility may exist to
E1 T
change the method and/or the procedure (box 13) in order to make the measuring process more cost
effective while increasing the uncertainty. A new iteration is then needed to estimate the resulting
measurement uncertainty, U (box 10);
E2
3) if U is not acceptable (i.e. if U > U ), the iteration process continues, or it is concluded that no
E1 E1 T
adequate measurement procedure is possible;
e) before the new iteration, analyze the relative magnitude of the uncertainty contributors. In many cases a few
uncertainty components pre-dominate the combined standard uncertainty and expanded uncertainty;
f) if U > U , then change the assumptions, the modelling or increase the knowledge about the uncertainty
E1 T
components (box 12) to make a more accurate (see 3.5 of VIM:1993) upper bound estimation of the largest
(dominant) uncertainty components;
g) make the second iteration of the uncertainty budget (boxes 7 to 9) leading to the second, lower and more
accurate (see 3.5 of VIM:1993) upper bound estimate of the uncertainty of measurement, U (box 10);
E2
h) compare the second estimated uncertainty U with the given target uncertainty, U (box A);
E2 T
1) if U is acceptable (i.e. if U u U ), then the uncertainty budget of the second iteration has proven that
E2 E2 T
the measurement procedure is adequate for the measuring task (box 11);
2) if U is not acceptable (i.e. if U > U ) then a third (and possibly more) iteration(s) is (are) needed.
E2 E2 T
Repeat the analysis of the uncertainty contributors (additional changes of assumptions, modelling and
increase in knowledge (box 12)) and concentrate on the currently largest uncertainty contributors;
i) when all possibilities has been used for making more accurate (lower) upper bound estimates of the measuring
uncertainties without coming to an acceptable measuring uncertainty U u U , then a change of the
EN T
measurement method or the measurement procedure or the conditions of measurement (box 13) is needed to
(possibly) bring down the magnitude of the estimated uncertainty, U . The iteration procedure starts again
EN
with a first iteration;
j) if changes in the measurement method or the measurement procedure or conditions (box 13) do not lead to an
acceptable uncertainty of measurement, the final possibility is to change the principle of measurement (box 14)
and start the above mentioned procedure again;
10 © ISO 1999 – All rights reserved

k) if change of the measuring principle and the related iterations described above do not lead to an acceptable
uncertainty of measurement the ultimate possibility is to change the measuring task and/or target uncertainty
(box 15) and start the above mentioned procedure again;
l) if change of measuring task or target uncertainty is not possible, it is demonstrated, that no adequate
measurement procedure exists (box 16).
© ISO 1999 – All rights reserved 11

Figure 2 — Procedure for Uncertainty of Measurement MAnagement (PUMA) for a measurement
process/procedure
12 © ISO 1999 – All rights reserved

7 Sources of errors and uncertainty of measurement
7.1 Types of errors
Different types of errors regularly shows up in measurement results.
� systematic errors;
� random errors;
� drift;
� outliers.
All errors are by nature systematic. When we see errors as non-systematic it is because the reason for the error is
not looked for or because the level of resolution is not sufficient. Systematic errors may be characterised by size
and sign (+ or�).
ER = MR� TV
where
ER is the error,
MR is the measurement result;
TV is thetruevalue.
Random errors are systematic errors caused by non-controlled random influence quantities. Random errors may be
characterized by the standard deviation and the type of distribution. The mean value of the random errors is often
considered as a basis for the evaluation of the systematic error (see Figure 3).
Key
1 Outlier
2 Dispersion 1
3 Dispersion 2
4 Systematic error 1
5 Systematic error 2
6 True value
Figure 3 — Types of errors in results of measurements
© ISO 1999 – All rights reserved 13

Drift is caused by a systematic influence of non-controlled influence quantities. Drift is often a time effect or a wear
effect. Drift may be characterized by change per unit time or per amount of use.
Outliers are caused by not repeatable incidents in the measurement. Noise — electrical or mechanical — may
result in outliers. A frequent reason for outliers is human mistakes as reading and writing errors or wrong handling
of measuring equipment. Outliers are impossible to characterize in advance.
Errors or uncertainties in a measuring process will be a mix of known and unknown errors from a number of
sources or error contributors.
The sources or contributors are not the same in each case, and the sum of the components are not the same.
It is still possible to make a systematic approach. There are always several sources or a combined effect of the ten
different ones indicated in Figure 4.
In the following, examples and further details about each of the ten contributors are given.
What is often difficult is that each of the contributors may act individually on the result of measurement. But in many
cases they even interfere with each other and cause additional errors and uncertainty.
Figure 4 and the following non-exhaustive lists (see 7.2 to 7.11) shall be used for getting ideas in a systematic way
when making uncertainty budgets. In each case the evaluation of the actual error/uncertainty component needs
knowledge about physics and/or experience in metrology.
In uncertainty budgets the uncertainty contributors and the uncertainty components may be grouped for
convenience.
Figure 4 —Uncertainty contributors in measurement
14 © ISO 1999 – All rights reserved

7.2 Environment for the measurement
In most cases — especially in GPS measurements — the temperature is the main uncertainty contributor of the
environment. Other uncertainty contributors may be:
� Temperature: absolute temperature, time � Gravity
variance, spatial gradient
� Electromagnetic interference
� Vibration/noise
� Transients in the power supply
� Humidity
� Pressured air (e.g. air bearings)
� Contamination
� Heat radiation
� Illumination
� Workpiece
� Ambient pressure
� Scale
� Air composition
� Instrument thermal equilibrium
� Air flow
7.3 Reference element of measurement equipment
The measuring equipment is divided into "reference element" and the "rest of the equipment", and it often pays to
look at the equipment that way.
� Stability � CCD-techniques
� Scale mark quality � Uncertainty of the calibration
� Temperature expansion coefficient � Resolution of the main scale (analogue or digital)
� Physical principle: line scale, optical digital scale, � Time since last calibration
magnetic digital scale, spindle, rack & pinion,
� Wavelength error
interferometer
7.4 Measurement equipment
� Interpretation system � Reading system
� Magnification, electrical or mechanical � Linear coefficient for thermal expansion
� Error wavelength � Temperature stability/sensitivity
� Zero-point stability � Parallaxes
� Force stability/absolute force � Time since last calibration
� Hysteresis � Response characteristic
� Guides/slideways � Interpolation system, error wavelength
� Probe system � Interpolation resolution
� Geometrical imperfections � Digitization
� Stiffness/rigidity
© ISO 1999 – All rights reserved 15

7.5 Measurement setup (excluding the placement and clamping of the workpiece)
In many cases there is no setup; the measurement equipment can measure "alone".
� Cosine errors and sine errors � Form deviation of tip
� Abbe principle � Stiffness of the probe system
� Temperature sensitivity � Optical aperture
� Stiffness/rigidity � Interaction between workpiece and setup
� Tip radius � Warming up
7.6 Software and calculations
Observe that even the number of digits or decimals can have an influence!
� Rounding/Quantification � Filtering
� Algorithms � Correction of algorithm/Certification of algorithm
� Implementation of algorithms � Interpolation/extrapolation
� Number of significant digits in the computation � Outlier handling
� Sampling
7.7 Metrologist
The human being is not stable; there is a difference from day to day and often a rather large change during the
day.
� Education � Knowledge (precision, appreciation)
� Experience � Honesty
� Training � Dedication
� Physical disadvantages/ability
7.8 Measurement object, workpiece or measuring instrument characteristic
� Surface roughness � Magnetism
� Form deviations � Hygroscopic characteristic of the material
� E-modulus (Young's modulus) � Ageing
� Stiffness beyond E-modulus � Cleanliness
� Temperature expansion coefficient � Temperature
� Conductivity � Internal stress
� Weight � Creep cha
...


SPÉCIFICATION ISO/TS
TECHNIQUE 14253-2
Première édition
1999-12-01
Spécification géométrique des produits
(GPS) — Vérification par la mesure des
pièces et des équipements de mesure —
Partie 2:
Guide pour l'estimation de l'incertitude
dans les mesures GPS, dans l’étalonnage
des équipements de mesure et dans la
vérification des produits
Geometrical product specifications (GPS) — Inspection by measurement
of workpieces and measuring equipment —
Part 2: Guide to the estimation of uncertainty in GPS measurement,
in calibration of measuring equipment and in product verification
Numéro de référence
©
ISO 1999
PDF – Exonération de responsabilité
Le présent fichier PDF peut contenir des polices de caractères intégrées. Conformément aux conditions de licence d'Adobe, ce fichier peut
être imprimé ou visualisé, mais ne doit pas être modifié à moins que l'ordinateur employé à cet effet ne bénéficie d'une licence autorisant
l'utilisation de ces polices et que celles-ci y soient installées. Lors du téléchargement de ce fichier, les parties concernées acceptent de fait la
responsabilité de ne pas enfreindre les conditions de licence d'Adobe. Le Secrétariat central de l'ISO décline toute responsabilité en la
matière.
Adobe est une marque déposée d'Adobe Systems Incorporated.
Les détails relatifs aux produits logiciels utilisés pour la création du présent fichier PDF sont disponibles dans la rubrique General Info du
fichier; les paramètres de création PDF ont été optimisés pour l'impression. Toutes les mesures ont été prises pour garantir l'exploitation de
ce fichier par les comités membres de l'ISO. Dans le cas peu probable où surviendrait un problème d'utilisation, veuillez en informer le
Secrétariat central à l'adresse donnée ci-dessous.
© ISO 1999
Droits de reproduction réservés. Sauf prescription différente, aucune partie de cette publication ne peut être reproduite ni utilisée sous quelque
forme que ce soit et par aucun procédé, électronique ou mécanique, y compris la photocopie et les microfilms, sans l'accord écrit de l’ISO à
l’adresse ci-après ou du comité membre de l’ISO dans le pays du demandeur.
ISO copyright office
Case postale 56 � CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax. + 41 22 734 10 79
E-mail copyright@iso.ch
Web www.iso.ch
ImpriméenSuisse
ii © ISO 1999 – Tous droits réservés

Sommaire Page
Avant-propos.iv
Introduction.v
1 Domaine d’application .1
2 Références normatives .2
3 Termes et définitions.2
4 Symboles.6
5 Concept de la méthode GUM itérative pour l'estimation de l'incertitude de mesure .7
6 Procédure pour le management de l'incertitude — PUMA.8
7 Sources d'erreurs et incertitude de mesure .13
8 Outils pour l'estimation des composantes d'incertitude, de l'incertitude-type et de l'incertitude
élargie .17
9 Estimation pratique de l'incertitude — Budgétisation de l'incertitude avec PUMA.26
10 Application .30
Annexe A (informative) Exemple de budgets d'incertitude — Étalonnage d'une bague de réglage.34
Annexe B (informative) Exemple de budgets d'incertitude — Conception d'une hiérarchie de
l'étalonnage .41
Annexe C (informative) Exemple de budgets d'incertitude — Mesure de circularité.66
Annexe D (informative) Relation avec la matrice GPS .72
Bibliographie .74
© ISO 1999 – Tous droits réservés iii

Avant-propos
L'ISO (Organisation internationale de normalisation) est une fédération mondiale d'organismes nationaux de
normalisation (comité membres de l'ISO). L'élaboration des Normes internationales est en général confiée aux
comités techniques de l'ISO. Chaque comité membre intéressé par une étude a le droit de faire partie du comité
technique créé à cet effet. Les organisations internationales, gouvernementales et non gouvernementales, en
liaison avec l'ISO, participent également aux travaux. L'ISO collabore étroitement avec la Commission
électrotechnique internationale (CEI) en ce qui concerne la normalisation électrotechnique.
Les Normes internationales sont rédigées conformément aux règles données dans les Directives ISO/CEI, Partie 3.
La tâche principale des comités techniques est d'élaborer les Normes internationales. Les projets de Normes
internationales adoptés par les comités techniques sont soumis aux comités membres pour vote. Leur publication
comme Normes internationales requiert l'approbation de 75 % au moins des comités membres votants.
Dans d'autres circonstances, en particulier lorsqu'il existe une demande urgente du marché, un comité technique
peut décider de publier d'autres types de documents normatifs:
— une Spécification publiquement disponible ISO (ISO/PAS) représente un accord entre les experts dans un
groupe de travail ISO et est acceptée pour publication si elle est approuvée par plus de 50 % des membres
votants du comité dont relève le goupe de travail;
— une Spécification technique ISO (ISO/TS) représente un accord entre les membres d'un comité technique et
est acceptée pour publication si elle est approuvée par plus de 2/3 des membres votants du comité.
Les ISO/PAS et ISO/TS font l'objet d'un nouvel examen tous les trois ans afin de décider éventuellement de leur
transformation en Normes internationales.
L'attention est appelée sur le fait que certains des éléments de la présente Spécification technique peuvent faire
l'objet de droits de propriété intellectuelle ou de droits analogues. L'ISO ne saurait être tenue pour responsable de
ne pas avoir identifié de tels droits de propriété et averti de leur existence.
L'ISO/TS 14253-2 a été élaborée par le comité technique ISO/TC 213, Spécifications et vérification
dimensionnelles et géométriques de produits.
L'ISO 14253 comprend les parties suivantes, présentées sous le titre général Spécification géométrique des
produits (GPS) — Vérification par la mesure des pièces et des équipements de mesure:
� Partie 1: Règles de décision pour prouver la conformité ou la non-conformité à la spécification
� Partie 2: Guide pour l’estimation de l’incertitude dans les mesures GPS, dans l’étalonnage des équipements
de mesure et dans la vérification des produits [Spécification technique]
� Partie 3: Procédures pour l'évaluation de l'intégrité des valeurs de l'incertitude de mesure
Les annexes A, B, C et D de la présente Spécification technique sont données uniquement à titre d'information.
iv © ISO 1999 – Tous droits réservés

Introduction
La présente Spécification technique est une spécification technique GPS globale (voir ISO/TR 14638:1995). Cette
Spécification technique GPS globale influence les maillons 4, 5 et 6 de toutes les chaînes de normes.
Pour des informations plus détaillées concernant la relation entre cette spécification, les autres normes et la
matrice GPS, voir l'annexe D.
La présente Spécification technique est développée pour venir à l'appui de l'ISO 14253-1. La présente
Spécification technique établit une procédure simplifiée et itérative du concept et de la façon d'évaluer et de
déterminer l'incertitude (incertitude-type et incertitude élargie) de mesure, et les recommandations pour
documenter et consigner les informations relatives à l'incertitude de mesure, telles qu'elles sont données dans le
«Guide pour l'expression de l'incertitude de mesure» (GUM). Dans la plupart des cas, des ressources très limitées
sont seulement nécessaires pour estimer une incertitude de mesure au moyen de cette procédure simplifiée et
itérative, mais cette dernière peut entraîner une légère surestimation de l'incertitude de mesure. Si une estimation
plus exacte de l'incertitude de mesure est nécessaire, les procédures plus élaborées du GUM doivent être
appliquées.
Cette procédure simplifiée et itérative des méthodes du GUM est destinée aux mesurages GPS, mais peut servir
dans d'autres domaines de la métrologie industrielle (appliquée).
L'incertitude de mesure et le concept de prise en compte de l'incertitude de mesure étant importants pour toutes
les fonctions techniques d'une société, la présente Spécification technique concerne par exemple, la fonction de
management, la fonction de conception et de développement, la fonction de fabrication, la fonction d'assurance
qualité, la fonction métrologie, etc.
La présente Spécification technique est particulièrement importante en ce qui concerne les systèmes d'assurance
qualité ISO 9000 dans lesquels il est exigé que l'incertitude de mesure soit connue [par exemple 4.11.1, 4.11.2 a)
et 4.11.2 b) de l'ISO 9001:1994].
Dans la présente Spécification technique, l'incertitude sur le résultat d'un processus d'étalonnage et d'un processus
de mesurage est abordée de la même façon:
� l'étalonnage est traité comme «une mesure de caractéristiques métrologiques d'un équipement de mesure ou
d'un étalon»;
� la mesure est traitée comme «une mesure de caractéristiques géométriques d'une pièce».
Par conséquent, dans la plupart des cas, il n'existe pas de distinction dans le texte entre mesurage et étalonnage.
Le terme «mesure» est utilisé comme synonyme des deux.
© ISO 1999 – Tous droits réservés v

SPÉCIFICATION TECHNIQUE ISO/TS 14253-2:1999(F)
Spécification géométrique des produits (GPS) — Vérification par
la mesure des pièces et des équipements de mesure —
Partie2:
Guide pour l'estimation de l'incertitude dans les mesures GPS,
dans l’étalonnage des équipements de mesure et dans la
vérification des produits
1 Domaine d’application
La présente Spécification technique donne des lignes directrices pour la mise en œuvre du concept de «Guide
pour l'estimation de l'incertitude de mesure» (en abrégé GUM), à appliquer dans l'industrie pour l'étalonnage
d'étalons et d'équipements de mesure dans le domaine GPS et la mesure des caractéristiques GPS de pièces.
L'objectif est de présenter des informations complètes sur la façon d'obtenir les causes d'incertitude et de fournir la
base d'une comparaison internationale des résultats de mesurages et de leurs incertitudes (relation entre le client
et le fournisseur).
La présente Spécification technique vient à l'appui de l'ISO 14253-1. Cette Spécification technique et l'ISO 14253-1
sont bénéfiques à toutes les fonctions techniques d'une société dans l'interprétation des spécifications GPS (à
savoir, les tolérances des caractéristiques d'une pièce et les valeurs des erreurs maximales tolérées (MPE:
Maximum Permissible Errors) pour les caractéristiques métrologiques de l'équipement de mesure).
La présente Spécification technique introduit la procédure pour le management de l'incertitude (PUMA: Procedure
for Uncertainty MAnagement), qui est une procédure pratique et itérative fondée sur le GUM pour estimer
l'incertitude de mesure sans modifier les concepts de base du GUM et qui est généralement destinée à être utilisée
pour estimer l'incertitude de mesure et pour donner des causes d'incertitude concernant:
� des résultats unitaires de mesure;
� la comparaison de deux résultats ou plus de mesure;
� la comparaison de résultats de mesure — à partir d'une ou de plusieurs pièces ou équipements de mesure —
avec des spécifications données (à savoir, les erreurs maximales tolérées (MPE) pour une caractéristique
métrologique d'un instrument de mesure ou un étalon, et les limites de tolérance pour une caractéristique de
pièce, etc.) pour prouver la conformité ou la non-conformité aux spécifications.
La méthode itérative est fondamentalement basée sur une stratégie de limite supérieure, à savoir la surestimation
de l'incertitude à tous les niveaux, mais les itérations déterminent la quantité de surestimation. Une surestimation
intentionnelle — et non une sous-estimation — est nécessaire pour empêcher la prise de mauvaises décisions sur
la base de résultats de mesure. La quantité de surestimation doit être contrôlée par l'évaluation économique de la
situation.
La méthode itérative est un outil pour maximiser les profits et réduire les coûts des activités métrologiques d'une
société. La méthode/procédure itérative est autorégulante sur le plan économique et est également un outil
permettant de modifier/réduire l'incertitude de mesure existante avec pour but de réduire le coût de la métrologie
(fabrication). La méthode itérative rend possible un compromis entre le risque, l'effort et le coût dans l'estimation et
la budgétisation de l'incertitude.
© ISO 1999 – Tous droits réservés 1

2 Références normatives
Les documents normatifs suivants contiennent des dispositions qui par suite de la référence qui y est faite,
constituent des dispositions valables pour la présente Spécification technique. Pour les références datées, les
amendements ultérieurs ou les révisions de ces publications ne s’appliquent pas. Toutefois, les parties prenantes
aux accords fondés sur la présente Spécification technique sont invitées à rechercher la possibilité d'appliquer les
éditions les plus récentes des documents normatifs indiqués ci-après. Pour les références non datées, la dernière
édition du document normatif en référence s’applique. Les membres de l'ISO et de la CEI possèdent le registre des
Normes internationales en vigueur.
ISO 1:1975, Température normale de référence des mesures industrielles de longueur.
ISO 4288:1996, Spécification géométrique des produits (GPS) — État de surface: Méthode du profil — Règles et
procédures pour l'évaluation de l'état de surface.
ISO 9001:1994, Systèmes qualité� Modèle pour l'assurance de la qualité en conception, développement,
production, installations et prestations associées.
ISO 9004-1:1994, Management de la qualité et éléments de systèmes qualité — Partie 1: Lignes directrices.
ISO 14253-1:1998, Spécification géométrique des produits (GPS) — Vérification par la mesure des pièces et
équipements de mesure� Partie 1: Règles de décision pour prouver la conformité ou la non-conformité à la
spécification.
1)
ISO 14253-3:— , Spécification géométrique des produits (GPS) — Vérification par la mesure des pièces et
équipements de mesure — Partie 3: Procédures pour l'évaluation de l'intégrité des valeurs de l'incertitude de
mesure.
ISO 14660-1:1999, Spécification géométrique des produits (GPS) — Termes généraux et définitions.
re
Guide pour l'expression de l'incertitude de mesure (GUM), 1 édition, 1995.
Vocabulaire international des termes fondamentaux et généraux de métrologie (VIM). BIPM, CEI, FICC, ISO,
e
OIML,UICPA,UIPPA, 2 édition, 1993.
3 Termes et définitions
Pour les besoins de la présente Spécification technique, les termes et définitions donnés dans l'ISO 14253-1,
l'ISO 14660-1, le VIM, le GUM et les suivants s'appliquent.
3.1
modèle de la boîte noire pour l'estimation de l'incertitude
méthode/modèle pour l'estimation de l'incertitude dans laquelle la valeur de sortie d'une mesure s'obtient dans la
même unité que l'entrée (stimuli), plutôt que par mesure d'autres grandeurs reliées de façon fonctionnelle au
mesurande
NOTE 1 Dans le modèle de la boîte noire — dans la présente Spécification technique — les composantes d'incertitude sont
supposées être additives, les grandeurs d'influence sont transformées en unité du mesurande et les coefficients de sensibilité
sont égaux à 1.
NOTE 2 Dans de nombreux cas, une méthode complexe de mesure peut être envisagée comme une simple boîte noire avec
un stimulus entrant et un résultat sortant de la boîte noire. Lorsqu'on ouvre une boîte noire, elle peut s'avérer contenir plusieurs
boîtes noires «plus petites» et/ou plusieurs boîtes transparentes.
1) À publier.
2 © ISO 1999 – Tous droits réservés

NOTE 3 La méthode d'estimation de l'incertitude reste une méthode de la boîte noire même s'il est nécessaire d'effectuer des
mesures supplémentaires pour déterminer les valeurs des grandeurs d'influence afin de réaliser les corrections
correspondantes.
3.2
modèle de la boîte transparente pour l'estimation de l'incertitude
méthode/modèle d'estimation d'incertitude dans laquelle la valeur d'un mesurande est obtenue par la mesure
d'autres grandeurs reliées de façon fonctionnelle au mesurande
3.3
opération de mesure
évaluation d'un mesurande selon sa définition
3.4
opération de mesure fondamentale (mesure fondamentale)
opération(s) de mesure qui constitue(nt) la base pour l'évaluation de caractéristiques plus compliquées d'une pièce
ou d'un équipement de mesure
NOTE Des exemples de mesure fondamentale sont:
a) une mesure parmi plusieurs mesures individuelles de l'écart de rectitude d'un élément d'une pièce;
b) une mesure parmi plusieurs mesures individuelles d'erreur d'indication d'un micromètre lors de la mesure de l'étendue de
l'erreur d'indication.
3.5
opération globale de mesure
opération compliquée de mesure qui est évaluée sur la base de plusieurs mesures fondamentales, éventuellement
différentes
NOTE Des exemples d'opération globale de mesure sont:
a) la mesure de la rectitude d'un élément d'une pièce;
b) l'étendue de l'erreur d'indication d'un micromètre.
3.6
incertitude élargie (d'une mesure)
U
[3.16 de l'ISO 14253-1:1998 et 2.3.5 du GUM:1995]
NOTE U (majuscule) indique toujours l'incertitude élargie de mesure.
3.7
incertitude vraie
U
A
incertitude de mesure qui serait obtenue par une estimation parfaite de l'incertitude
NOTE 1 Les incertitudes vraies sont, par nature, indéterminées.
NOTE 2 Voir également 8.8.
3.8
incertitude conventionnellement vraie — Incertitude GUM
U
C
incertitude de mesure estimée entièrement selon les procédures les plus élaborées du GUM
NOTE 1 L'incertitude conventionnellement vraie de mesure peut différer d'une incertitude de mesure estimée conformément
à la présente Spécification technique.
NOTE 2 Voir également 8.8.
© ISO 1999 – Tous droits réservés 3

3.9
incertitude approchée
U
EN
incertitude de mesure estimée par la méthode simplifiée et itérative
NOTE 1 L'indice N indique que U est estimée par le nombre d'itérations N. La désignation U peut être utilisée sans
EN E
indication du nombre d'itérations, lorsqu'il n'est pas important de connaître le nombre d'itérations.
NOTE 2 Voir également 8.8.
3.10
incertitude cible (pour une mesure ou un étalonnage)
U
T
incertitude déterminée comme étant l'optimum pour l'opération de mesure
NOTE 1 L'incertitude cible est le résultat d'une décision de direction impliquant par exemple, la conception, la fabrication, le
service d'assurance qualité, la commercialisation, les ventes et la distribution.
NOTE 2 L'incertitude cible est déterminée (optimisée) en tenant compte de la spécification [tolérance ou erreur maximale
tolérée (MPE)], de l'aptitude du processus, des coûts, de la criticité et des exigences de 4.11.1 et 4.11.2 de l'ISO 9001:1994,
13.1 de l'ISO 9004:1994 et de l'ISO 14253-1.
NOTE 3 Voir également 8.8.
3.11
incertitude requise de mesure
U
R
incertitude requise pour un processus et une opération donnés de mesure
NOTE Voir également 6.2. L'incertitude requise peut être spécifiée par un client, par exemple.
3.12
management de l'incertitude
processus consistant à dériver un mode opératoire de mesure adéquat à partir d'une opération de mesure et de
l'incertitude cible en utilisant des techniques de budgétisation de l'incertitude
3.13
budget d'incertitude (pour une mesure ou un étalonnage)
déclaration résumant l'estimation des composantes d'incertitude qui contribuent à l'incertitude d'un résultat de
mesure
NOTE 1 L'incertitude du résultat de la mesure n'est pas ambiguë uniquement lorsque le mode opératoire de mesure (y
compris l'objet de mesure, le mesurande, la méthode et les conditions de mesure) est défini.
NOTE 2 Le terme «budget» est utilisé pour l'attribution de valeurs numériques aux composantes d'incertitude, à leur
combinaison et leur élargissement, sur la base du mode opératoire de mesure, des conditions et hypothèses de mesure.
3.14
cause d'incertitude
xx
source d'incertitude de mesure pour un processus de mesure
3.15
valeur limite (limite d'écart) pour une cause d'incertitude
a
xx
valeur absolue de la (ou des) valeur(s) extrême(s) de la cause d'incertitude, xx
4 © ISO 1999 – Tous droits réservés

3.16
composante d'incertitude
u
xx
incertitude-type de la cause d'incertitude, xx
NOTE La méthode d'itération utilise la désignation u pour toutes les composantes d'incertitude. Cela n'est pas cohérent
xx
avec la version actuelle du GUM qui utilise parfois la désignation s pour les composantes d'incertitude évaluées par
xx
l'évaluation A et la désignation u pour les composantes d'incertitude évaluées par l'évaluation B.
xx
3.17
grandeur d'influence d'un instrument de mesure
caractéristique d'un instrument de mesure qui affecte le résultat d'une mesure effectuée par l'instrument
3.18
grandeur d'influence d'une pièce
caractéristique d'une pièce qui affecte le résultat d'une mesure effectuée sur cette pièce
© ISO 1999 – Tous droits réservés 5

4 Symboles
Pour les besoins de la présente Spécification technique, les symboles génériques du Tableau 1 s'appliquent.
Tableau 1 — Symboles génériques
Symbole Description
a valeur limite pour une distribution
a valeur limite pour une erreur ou une cause d'incertitude (dans l'unité du résultat de mesure — du mesurande)
xx
a* valeur limite pour une erreur ou une cause d'incertitude (dans l'unité de la grandeur d'influence)
xx
� coefficient de dilatation thermique linéaire
b coefficient pour la transformation de a en u
xx xx
C correction (valeur)
d résolution d'un équipement de mesure
E module d'Young
ER erreur (valeur d'une mesure)
fonction de plusieurs valeurs de mesure [G(X , X , ., X ,.)]
G
1 2 i
h valeur d'hystérésis
k facteur d'élargissement
m nombre d'écarts-types dans la moitié d'un intervalle de confiance
MR résultat de mesure (valeur)
n nombre de .
N nombre d'itérations
� nombre de Poisson
p nombre de causes d'incertitude totale non corrélées
r nombre de causes d'incertitude totale corrélées
coefficient de corrélation

TV valeur vraie d'une mesure
u, u incertitude-type (écart-type)
i
s
écart-type d'un échantillon
x
s écart-type d'une valeur moyenne d'un échantillon
x
u incertitude-type composée
c
u écart-type de la cause d'incertitude xx — composante d'incertitude
xx
U incertitude élargie de mesure
U incertitude vraie de mesure
A
U
incertitude conventionnellement vraie de mesure
C
U incertitude approchée d'une mesure (nombre d'itérations non indiqué)
E
U
incertitude approchée d'une mesure du nombre d'itérations N
EN
U incertitude requise
R
U
incertitude cible
T
U valeur d'incertitude (non estimée selon le GUM ou la présente Spécification technique)
V
X résultat de mesure (brut)
X résultat de mesure (dans le modèle de la boîte transparente d'estimation de l'incertitude)
i
Y résultat de mesure (corrigé)
6 © ISO 1999 – Tous droits réservés

5 Concept de la méthode GUM itérative pour l'estimation de l'incertitude de mesure
En appliquant entièrement la méthode GUM, on trouve une incertitude conventionnellement vraie de mesure, U .
C
La méthode/mode opératoire simplifié et itératif de la présente Spécification technique permet d'obtenir des
incertitudes estimées de mesure, U , en surestimant les composantes/causes d'influence de l'incertitude
E
(U W U ). Le processus de surestimation permet de prendre en compte le «pire des cas» à la limite supérieure de
E C
chaque cause d'incertitude connue ou prévisible, ce qui assure des résultats d'estimations «par sécurité», c'est-à-
dire sans sous-estimation de l'incertitude de mesure. La méthode simplifiée et itérative de la présente Spécification
technique est fondée sur ce qui suit:
� toutes les causes d'incertitude sont identifiées;
� il est décidé des éventuelles corrections qui doivent être effectuées (voir 8.4.6);
� l'influence de l'incertitude du résultat de mesure à partir de chaque cause est évaluée sous forme d'incertitude-
type u , dénommée la composante d'incertitude;
xx
NOTE Par convention dans la méthode itérative, l'influence de chaque cause doit être convertie dans l'unité du mesurande
au moyen des équations/formules physiques et des coefficients de sensibilité applicables.
� un processus d'itération, PUMA (voir article 6);
� l'évaluation de chacune des composantes d'incertitude (incertitudes types) u peut prendre place soit par
xx
l'évaluation de type A, soit par l'évaluation de type B;
� l'évaluation de type B est préférable — si possible — dans la première itération de façon à obtenir une
estimation grossière de l'incertitude pour établir un aperçu et économiser des coûts;
� l'effet total de toutes les causes (dénommé incertitude-type composée) est calculé au moyen de la formule:
2 222
uu=+u+u+.+u (1)
c1xx2 x3 xn
� la formule (1) n'est valable que pour un modèle de la boîte noire d'estimation de l'incertitude et lorsque les
composantes u sont toutes non corrélées (pour plus de détails et d'autres formules, voir 8.6 et 8.7);
xx
� pour simplifier, les seuls coefficients de corrélation entre les causes concernées sont
� =1,� 1, 0 (2)
� si la corrélation des composantes d'incertitude n'est pas connue, une corrélation complète est supposée, � =1
ou � 1. Les composantes corrélées sont additionnées arithmétiquement avant d'être insérées dans la formule
ci-dessus (voir 8.5 et 8.6);
� l'incertitude élargie U est calculée au moyen de la formule:
U = k � u (3)
c
où k =2; k est le facteur d'élargissement (voir également 8.8);
� la méthode simplifiée et itérative consiste généralement en au moins deux itérations de l'estimation des
composantes d'incertitude.
� La première itération très grossière, rapide et bon marché a pour objet d'identifier les composantes les plus
importantes de l'incertitude (voir Figure 1).
© ISO 1999 – Tous droits réservés 7

� Les itérations suivantes, le cas échéant, ne consistent qu'à effectuer des estimations de «limite supérieure»
plus exactes des plus importantes composantes pour abaisser l'estimation de l'incertitude (u et U)à un
c
éventuel ordre de grandeur acceptable.
La méthode simplifiée et itérative peut être utilisée pour deux buts:
a) le management de l'incertitude de mesure pour le résultat d'un processus de mesure donné (peut servir aux
résultats à partir d'un processus connu de mesure ou pour comparaison de deux ou plusieurs de ces
résultats), voir 6.2;
b) le management de l'incertitude pour un processus de mesure. Développement d'un processus de mesure
adéquat, à savoir U u U (voir 6.3).
E T
6 Procédure pour le management de l'incertitude — PUMA
6.1 Généralités
La condition préalable à la budgétisation et au management de l'incertitude est une opération de mesure
clairement identifiée et définie; à savoir, le mesurande à quantifier (une caractéristique GPS d'une pièce ou une
caractéristique métrologique d'un équipement de mesure de GPS). L'incertitude de mesure est une mesure de la
qualité de la valeur mesurée selon les définitions d'une caractéristique GPS de la pièce ou une caractéristique
métrologique de l'équipement de mesure GPS donné dans les normes GPS.
Les normes GPS définissent les «valeurs conventionnellement vraies» [voir 1.20 du VIM (1993)] des
caractéristiques à mesurer par des chaînes de normes et des normes globales (voir l'ISO/TR 14638). Très
souvent, les normes GPS définissent également le principe de mesure idéal — ou conventionnellement vrai — [voir
2.3 du VIM (1993)], la méthode de mesure [voir 2.4 du VIM (1993)], le mode opératoire de mesure [voir 2.5 du VIM
(1993)] et les «conditions de référence» [voir 5.7 du VIM (1993)].
Les écarts par rapport aux valeurs normalisées conventionnellement vraies des caractéristiques, etc. (l'opérateur
idéal) contribuent à l'incertitude de mesure.
6.2 Management de l'incertitude pour un processus donné de mesure
Le management de l'incertitude de mesure pour une opération de mesure donnée (case 1 de la Figure 1) et pour
un processus de mesure existant, est illustré à la Figure 1. Le principe de mesure (case 3), la méthode de mesure
(case 4), le mode opératoire de mesure (case 5) et les conditions de mesure (case 6) sont fixés et donnés ou
décidés dans ce cas, et ne peuvent être modifiés. La seule tâche consiste à évaluer la conséquence sur
l'incertitude de mesure. Un U requis peut être donné ou décidé.
R
En utilisant la méthode GUM itérative, la première itération ne sert qu'à l'orientation et à la recherche des causes
dominantes d'incertitude. La seule chose à faire, dans le processus de management dans ce cas, est d'affiner
l'estimation des causes dominantes afin d'approcher une estimation vraie des composantes d'incertitude, en
évitant ainsi une surestimation trop importante, si nécessaire.
8 © ISO 1999 – Tous droits réservés

Figure 1 — Management de l'incertitude pour un résultat de mesure à partir d'un processus de mesure
donné
La procédure est la suivante:
a) effectuer la première itération de préférence sur la base d'un modèle de boîte noire du processus d'estimation
d'incertitude et établir un budget d'incertitude préliminaire (cases 7 à 9) aboutissant à la première estimation
grossière de l'incertitude élargie, U (case 10). Pour des détails concernant l'estimation de l'incertitude, voir
E1
article 9. Toutes les estimations des incertitudes U s'effectuent sous la forme d'estimations de limite
EN
supérieure;
b) comparer la première incertitude estimée, U , avec l'incertitude requise U (case A) pour l'opération de
E1 R
mesure réelle;
1) si U est acceptable (c'est-à-dire si U u U ), alors le budget d'incertitude de la première itération a
E1 E1 R
prouvé que le mode opératoire de mesure donné est adapté à l'opération de mesure (case 11);
2) si U n'est pas acceptable (c'est-à-dire si U � U ), ou s'il n'y a pas d'incertitude requise, mais qu'une
E1 E1 R
valeur inférieure et plus vraie est souhaitée, le processus d'itération se poursuit;
c) avant la nouvelle itération, analyser l'ordre de grandeur relatif des causes d'incertitude. Dans de nombreux
cas, quelques composantes d'incertitude dominent l'incertitude-type composée et l'incertitude élargie;
d) modifier les hypothèses ou améliorer les connaissances au sujet des composantes d'incertitude afin
d'effectuer une estimation de limite supérieure plus exacte [voir 3.5 du VIM (1993)] des composantes
d'incertitude les plus importantes (dominantes) (case 12);
modifier pour un modèle plus détaillé du processus d'estimation d'incertitude ou une résolution supérieure du
processus de mesure (case 12);
e) effectuer la seconde itération du budget d'incertitude (cases 7 à 9) aboutissant à la seconde estimation,
inférieure et plus exacte [voir 3.5 du VIM (1993)], de limite supérieure de la mesure d'incertitude, U (case
E2
10);
f) comparer la seconde incertitude estimée, U , (case A) avec l'incertitude requise, U , pour l'opération de
E2 R
mesure réelle;
1) si U est acceptable (c'est-à-dire si U u U ), alors le budget d'incertitude de la seconde itération a
E2 E2 R
prouvé que le mode opératoire de mesure donné est adapté à l'opération de mesure (case 11);
2) si U n'est pas acceptable (c'est-à-dire si U � U ), ou s'il n'y a pas d'incertitude requise, mais qu'une
E2 E2 R
valeur inférieure et plus exacte est souhaitée, alors une troisième (et éventuellement d'autres) itération(s)
est (sont) nécessaire(s). Répéter l'analyse des causes d'incertitude [modifications additionnelles des
hypothèses, amélioration des connaissances, modifications de la modélisation, etc. (case 12)], et se
concentrer sur les causes d'incertitude les plus importantes actuellement;
© ISO 1999 – Tous droits réservés 9

g) lorsque toutes les possibilités ont été utilisées pour effectuer des estimations de limite supérieure plus exactes
(inférieures) des incertitudes de mesure sans arriver à une incertitude de mesure acceptable U u U ,alors
EN R
il est prouvé qu'il n'est pas possible de satisfaire à l'exigence donnée U .
R
6.3 Management de l'incertitude pour la conception et le développement d'un processus/mode
opératoire de mesure
Le management de l'incertitude dans ce cas est effectué pour développer un mode opératoire de mesure adéquat
[mesure des caractéristiques géométriques d'une pièce ou des caractéristiques métrologiques d'un équipement de
mesure (étalonnage)]. Le management de l'incertitude s'effectue sur la base d'une opération de mesure définie
(case 1 de la Figure 2) et sur une incertitude cible donnée, U (case 2 de la Figure 2). La définition de l'opération
T
de mesure et de l'incertitude cible sont des décisions de la société qui doivent être prises à un niveau de direction
suffisamment élevé. Un mode opératoire de mesure adéquat est un mode opératoire qui aboutit à une incertitude
estimée de mesure inférieure ou égale à l'incertitude cible. Si l'incertitude de mesure estimée est bien inférieure à
l'incertitude cible, le mode opératoire de mesure peut ne pas être (économiquement) optimal pour exécuter
l'opération de mesure (à savoir, le processus de mesure est trop onéreux).
Le PUMA, fondé sur une opération de mesure donnée (case 1) et une incertitude cible donnée U (case 2)
T
comporte ce qui suit (voir Figure 2):
a) choisir le principe de mesure (case 3), sur la base de l'expérience et des éventuels instruments de mesure
présents dans la société;
b) établir et documenter une méthode de mesure préliminaire (case 4), un mode opératoire de mesure (case 5)
et des conditions de mesure (case 6), sur la base de l'expérience et des possibilités connues dans la société;
c) effectuer la première itération de préférence sur la base d'un modèle de boîte noire du processus d'estimation
de l'incertitude et établir un budget préliminaire d'incertitude (cases 7 à 9) aboutissant à la première estimation
grossière de l'incertitude élargie, U (case 10). Pour des détails concernant l'estimation de l'incertitude, voir
E1
l'article 9. Toutes les estimations des incertitudes U s'effectuent sous la forme d'estimations de limite
EN
supérieure;
d) comparer la première incertitude estimée, U , avec l'incertitude cible donnée U (case A):
E1 T
1) si U est acceptable (c'est-à-dire si U u U ), alors le budget d'incertitude de la première itération a
E1 E1 T
prouvé que le mode opératoire de mesure est adapté à l'opération de mesure (case 11);
2) si U << U , alors le mode opératoire de mesure est techniquement acceptable mais il peut exister une
E1 T
possibilité de modifier la méthode et/ou le mode opératoire (case 13) afin de rendre le processus de
mesure plus rentable, tout en augmentant l'incertitude. Une nouvelle itération est alors nécessaire pour
estimer l'incertitude de mesure qui en résulte, U (case 10);
E2
3) si U n'est pas acceptable (c'est-à-dire si U > U ), le processus d'itération se poursuit, ou il est conclu
E1 E1 T
qu'aucun mode opératoire de mesure adéquat n'est possible;
e) avant la nouvelle itération, analyser l'ordre de grandeur relatif des causes d'incertitude; dans de nombreux
cas, quelques composantes d'incertitude prédominent l'incertitude-type composée et l'incertitude élargie;
f) si U > U , alors modifier les hypothèses, la modélisation ou améliorer les connaissances au sujet des
E1 T
composantes d'incertitude (case 12) afin d'effectuer une estimation plus exacte [voir 3.5 du VIM (1993)] de
limite supérieure des composantes d'incertitude les plus importantes (dominantes);
g) effectuer la seconde itération du budget d'incertitude (cases 7 à 9) aboutissant à la seconde estimation de
limite supérieure, inférieure et plus exacte [voir 3.5 du VIM (1993)] de la mesure d'incertitude, U (case 10);
E2
10 © ISO 1999 – Tous droits réservés

h) comparer la seconde incertitude estimée, U , (case A) avec l'incertitude cible donnée, U (case A);
E2 T
1) si U est acceptable (c'est-à-dire si U u U ), alors le budget d'incertitude de la seconde itération a
E2 E2 T
prouvé que le mode opératoire de mesure donné est adapté à l'opération de mesure (case 11);
2) si U n'est pas acceptable (c'est-à-dire si U � U ), alors une troisième (et éventuellement d'autres)
E2 E2 T
itération(s) est (sont) nécessaire(s). Répéter l'analyse des causes d'incertitude [modifications supplémen-
taires des hypothèses, amélioration des connaissances, modifications de la modélisation, etc. (case 12)],
et se concentrer sur les causes d'incertitude courantes les plus importantes;
i) lorsque toutes les possibilités ont été utilisées pour effectuer des estimations de limite supérieure plus exactes
(inférieures) des incertitudes de mesure sans arriver à une incertitude de mesure acceptable U u U ,alors
EN T
une modification de la méthode de mesure ou du mode opératoire de mesure ou des conditions de mesure
(case 13) est nécessaire pour (éventuellement) réduire l'ordre de grandeur de l'incertitude estimée, U .Le
EN
mode opératoire d'itération recommence avec une première itération;
j) si des modifications dans la méthode de mesure ou le mode opératoire ou les conditions de mesure (case 13)
n'aboutissent pas à une incertitude de mesure acceptable, la dernière possibilité consiste à modifier le principe
de mesure (case 14) et de recommencer le mode opératoire au début;
k) si la modification du principe de mesure et des itérations liées décrites ci-dessus n'aboutit pas à une incertitude
de mesure acceptable, la toute dernière possibilité est de modifier l'opération de mesure et/ou l'incertitude
cible (case 15) et de recommencer le mode opératoire ci-dessus mentionné;
l) si la modification de l'opération de mesure ou de l'incertitude cible n'est pas possible, il est démontré qu'aucun
mode opératoire de mesure adéquat n'existe (case 16).
© ISO 1999 – Tous droits réservés 11

Figure 2 —Procédure pour le management de l'incertitude de mesure (PUMA) pour un processus/mode
opératoiredemesure
12 © ISO 1999 – Tous droits réservés

7 Sources d'erreurs et incertitude de mesure
7.1 Types d'erreurs
Différents types d'erreurs se produisent régulièrement dans les résultats de mesure:
� erreurs systématiques;
� erreurs aléatoires;
� dérives;
� valeurs aberrantes.
Toutes les erreurs sont systématiques par nature. Lorsque des erreurs sont considérées comme n'étant pas
systématiques, c'est que la cause de l'erreur n'est pas recherchée ou que le niveau de résolution n'est pas
suffisant. Les erreurs systématiques peuvent être caractérisées par la taille et le signe (+ ou –).
ER = MR � TV (4)

ER est l'erreur;
MR est le résultat de mesure;
TV est la valeur vraie.
Les erreurs aléatoires sont des erreurs systématiques causées par des grandeurs d'influence aléatoires non
contrôlées. Les erreurs aléatoires peuvent être caractérisées par l'écart-type et le type de distribution. La valeur
moyenne des erreurs aléatoires est souvent considérée comme une base d'évaluation de l'erreur systématique
(voir Figure 3).
Légende
1 Valeur aberrante
2 Dispersion 1
3 Dispersion 2
4 Erreur systématique 1
5 Erreur systématique 2
6 Valeur vraie
Figure 3 — Types d'erreurs dans les résultats de mesure
© ISO 1999 – Tous droits réservés 13

La dérive est causée par une influence systématique de grandeurs d'influence non contrôlées. La dérive est
souvent un effet temporel ou un effet d'usure. La dérive peut être caractérisée par la modification par unité de
temps ou par quantité d'utilisation.
Les valeurs aberrantes sont causées par des incidents non répétables dans la mesure. Le bruit, électrique ou
mécanique, peut entraîner des valeurs aberrantes. Une raison fréquente des valeurs aberrantes est l'erreur
humaine dans la lecture ou l'écriture d'erreurs ou une mauvaise manipulation des équipements de mesure. Les
valeurs aberrantes sont impossibles à caractériser à l'avance.
Les erreurs ou incertitudes dans un processus de mesure sont un mélange d'erreurs connues et inconnues à partir
d'un certain nombre de sources ou de causes d'erreur.
Les sources ou causes ne sont pas les mêmes dans chaque cas, et la somme des composantes n'est pas la
même.
Il est possible d'effectuer une approche systématique. Il existe toujours plusieurs sources ou un effet combiné des
dix sources différentes indiquées à la Figure 4.
Dans ce qui suit, des exemples et d'autres détails concernant chacune des dix causes sont donnés.
Souvent, la difficulté est que chacune des causes peut agir individuellement sur le résultat de mesure. Mais dans
de nombreux cas, elles interfèrent les unes avec les autres et entraînent des erreurs et une incertitude
supplémentaires.
La Figure 4 et les listes non exhaustives suivantes (voir 7.2 à 7.11) doivent servir de façon systématique pour
obtenir des idées lors de l'établissement des budgets d'incertitude. Dans chaque cas, l'évaluation de la
composante réelle erreur/incertitude nécessite des connaissances en physique et/ou une expérience en
métrologie.
Dans les budgets d'incertitude, les causes d'incertitude et les composantes d'incertitude peuvent être regroupées
pour des raisons pratiques.
Figure 4 — Causes d'incertitude dans la mesure
14 © ISO 1999 – Tous droits réservés

7.2 Environnement pour la mesure
Dans la plupart des cas, en particulier dans les mesures GPS, la température est la principale cause d'incertitude
d'environnement. D'autres causes d'incertitude peuvent être:
— Température: température absolue, gradient —Gravité
temporel, gradient spatial
— Interférence électromagnétique
— Vibration/bruit
— Transitoires dans les sources d'alimentation
— Humidité
— Air comprimé (par exemple paliers
—Pollution pneumatiques)
— Éclairage — Rayonnement thermique
— Pression ambiante —Pièce
— Composition de l'air — Échelle (graduation)
— Flux d'air — Équilibre thermique de l'instrument
7.3 Élément de référence de l'équipement de mesure
L'équipement de mesure comporte «l'élément de référence» et le «reste de l'équipement», et il est souvent
intéressant de considérer l'équipement de cette façon.
— Stabilité — Techniques CCD
— Qualité des repères — Incertitude de l'étalonnage
— Coefficient de dilatation thermique — Résolution de l'affichage principal (analogique ou
numérique)
— Principe physique de fonctionnement: règle à
traits, règle optique numérique, règle numérique — Durée depuis le dernier étalonnage
magnét
...

Questions, Comments and Discussion

Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.

Loading comments...