Edgar DietrichAlfred Schulze

Measurement Process QualificationGage Acceptance and Measurement Uncertainty According to Current Standards

Edgar Dietrich /Alfred Schulze

Measurement Process Qualification

MeasurementProcess Qualification

Edgar DietrichAlfred Schulze

Gage Acceptance and Measurement UncertaintyAccording to Current Standards

Hanser Publishers, Munich Hanser Publications, Cincinnati

© 2011 Carl Hanser Verlag, Munichwww.hanser.deLecturer: Dipl.-Ing.Volker HerzbergProduction Management: Der Buchmacher, Arthur Lenner, MünchenTypesetting: Heide MesadCoverconcept: Marc Müller-Bremer, www.rebranding.de, München, GermanyCoverillustration: Atelier Frank Wohlgemuth, BremenCoverrealisation: Stephan RönigkPrinted and bound by Druckhaus »Thomas Müntzer« GmbH, Bad Langensalza, GermanyPrinted in Germany

The authors:Dr.-Ing. Edgar Dietrich, Dipl.-Ing. Alfred Schulze, Q-DAS GmbH & Co. KG, Weinheim.

Distributed in the USA and in Canada byHanser Publications6915 Valley Avenue, Cincinnati, Ohio 45244-3029, USAFax: (513) 527-8801Phone: (513) 527-8896 or 1-800-950-8977www.hanserpublications.com

Distributed in all other countries byCarl Hanser VerlagPostfach 86 04 20, 81631 München, GermanyFax: +49 (89) 98 48 09www.hanser.de

The use of general descriptive names, trademarks, etc., in this publication, even if the former are notespecially identified, is not to be taken as a sign that such names, as understood by the Trade Marksand Merchandise Marks Act, may accordingly be used freely by anyone.While the advice and information in this book are believed to be true and accurate at the dateof going to press, neither the authors nor the editors nor the publisher can accept any legalresponsibility for any errors or omissions that may be made. The publisher makes no warranty,express or implied, with respect to the material contained herein.

Library of Congress Cataloging-in-Publication Data

Dietrich, Edgar, 1951-[Eignungsnachweis von Pruffprozessen. English]Measurement process qualification : gage acceptance and measurement

uncertainty according to current standards / Edgar Dietrich, Alfred Schulze.p. cm.

Includes bibliographical references and index.ISBN-13: 978-1-56990-505-0 (hardcover)ISBN-10: 1-56990-505-3 (hardcover)ISBN-13: 978-3-446-42407-4 (hardcover)1. Measurement. 2. Tolerance (Engineering) 3. Gages. I. Schulze,

Alfred, 1952- II. Title.T50.D53613 2011620‘.0044--dc22

2011010487

Bibliografische Information Der Deutschen BibliothekDie Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie;detaillierte bibliografische Daten sind im Internet über <http://dnb.d-nb.de> abrufbar.

German ISBN: 978-3-446-42407-4US-ISBN: 978-1-56990-505-0E-Book-ISBN: 978-3-446-42955-0

All rights reserved. No part of this book may be reproduced or transmitted in any form or byany means, electronic or mechanical, including photocopying or by any information storage andretrieval system, without permission in writing from the publisher.

v

Preface

Measurement process capability studies gained in importance during the last years. Atthe end of the 1980s, there were only few company guidelines referring to thesignificance of gages and only some of them required capability studies. Over the years,new and further guidelines were added. The procedure was more and more refined andits application improved. After the methodology had established itself, more and morerequirements were added from quasi-standards, such as the QS-9000 or VDA 6.1guidelines. From now on, measurement process capability studies have to beconducted regularly in order to obtain the respective QM system certificate.

Today, the field of capability studies also includes the determination of measurementuncertainties that is to be applied in production. As an example, ISO 14253 [29] requiresthe determination of the measurement uncertainty for measures of length. Thisuncertainty must be taken into account at the specification limits. Hence, more andmore companies are obliged to include the calculation of the measurement uncertaintyin their QM system and to administer it in the respective application. In order to facilitatethese procedures as much as possible, the German Association of the AutomotiveIndustry (VDA) has already published the “Measurement Process Capability” guideline(VDA 5 [70]). As the title indicates, this guide does not only deal with measuringdevices. It is also about all the influencing factors affecting the measuring device.

The second and third edition of our book “Statistical Procedures for Machine andProcess Qualification” [13] contains a chapter about “gage capability”. Due to the varietyof procedures for measurement process capability studies, we decided to leave out thistopic in the fourth edition and to publish this book instead. In order to better define thissubject in detail.

Special thanks go to Mr. Ofen (Robert Bosch GmbH, Bamberg) for the long-termcooperation and his professional support. Major parts of the book “Sonderfälle bei derBeurteilung von Messverfahren” (special cases in the evaluation of measurementprocedures) [67] are penned by him. Upon his approval, we adopted some passages inthis book.

Our thanks also go to Ms. Mesad for the layout and the textual and graphicalpresentation of this book.

In case of questions, please contact us directly. Q-DAS® GmbH, Eisleber Str. 2, 69469Weinheim, Germany, phone: +49 6201 3941 0, fax: +49 6201 3941 24, hotline: +496201 3941 14, e-mail: q-das@q-das.de.

Weinheim, April 2003Edgar Dietrich and Alfred Schulze

vi

Preface to the 2nd Edition

The subject this book received a great many response. We received a significantamount of feedback on the first edition including many suggestions and proposals. Inparticular, our thanks goes to those readers who advised us of some inconsistencies.We acted on these suggestions and the new edition considers that helpful input.

By now, VDA 5 “Measurement Process Capability” has been published. It caused manydiscussions among experts but also raised many questions. Hence, we expanded onthis particular subject and added some sample calculations.

In many conversations and seminars, people have always expressed the wish to have adifferent procedure for the determination of the “extended measurement uncertainty”.This procedure is to be structured in a similar way as the widely-used “R&R” procedurefor the determination of gage capability. So we developed AIO procedure (all-in-onemethod) for the determination of extended measurement uncertainty. This procedure isbased on current draft standards and facilitates the determination of individual standarduncertainties in a step by step proceedure. The final result is the calculated extendedmeasurement uncertainty.

Graduate engineer Michael Radeck supported us in providing new sample calculations.He also edited the “attribute gage” subject. We would like to thank him for hisassistance.

Weinheim, Mai 2004Edgar Dietrich and Alfred Schulze

Preface to the 3rd Edition

The second edition already contained the determination of the measurement uncertaintyaccording to the “Guide to the Uncertainty in Measurement” (GUM [32]) but itsapplication was uncommon in industrial production. However, things have changedlately.

Particularly due to the new ISO 10012:2004 [29] “Measurement Management Systems -Requirements for Measurement Processes and Measuring Equipment” standard, thissubject has gained in importance. As the standard’s subheading indicates, themeasurement uncertainty must be determined for the respective measurement process.The standard says: “The measurement uncertainty must be estimated for everymeasurement process that is monitored by the measurement system. The Guide to theExpression of Uncertainty in Measurement (GUM) [32] contains the methods to beapplied.”

Due to this fact, we decided to deal with measurement uncertainty in more detail in thisbook.

vii

Since the publication of the second edition in May 2004, further company guidelinesabout “measurement process capability studies” have been launched. We added theDaimlerChrysler LF05 guideline and the Robert Bosch GmbH booklet 10 in this edition.Both guidelines implemented the procedures and methods for measurement processcapability studies that this book describes in theory. Today, companies and suppliershave gained experience in using these methods by applying these guidelines. Thepractical benefit of the determination of measurement uncertainty is confirmed.

Graduate engineer Stephan Conrad supported us in writing the “Determination ofMeasurement Uncertainty” chapter. We would like to thank him for his assistance.

In case of questions, please contact us directly. Q-DAS® GmbH, Eisleber Str. 2, 69469Weinheim, Germany, phone: +49 6201 3941 0, fax: +49 6201 3941 24, hotline: +496201 3941 14, e-mail: q-das@q-das.de.

Weinheim, September 2006Edgar Dietrich and Alfred Schulze

ix

Table of Contents

Preface .......................................................................................................................... vPreface to the 2nd Edition................................................................................................ viPreface to the 3rd Edition ................................................................................................ viTable of Contents ........................................................................................................... ix1 Measurement Process Capability..............................................................................1

1.1 Introduction ...................................................................................................................11.1.1 Why Measurement Process Capability? .........................................................1

1.2 Historical Retrospect and Prospect...............................................................................71.2.1 Development “Measurement Process Capability” ...........................................9

1.3 Notes from the Authors about MSA [1] and VDA 5 [70] ..............................................111.4 Experimental Evaluation .............................................................................................12

2 Gage Monitoring as a Basis for Measurement Process Capability..........................162.1 Gage Calibration .........................................................................................................162.2 Dial Gage Calibration ..................................................................................................172.3 Capability Studies for Standard Gages .......................................................................19

3 Definitions and Terms .............................................................................................223.1 Process .......................................................................................................................223.2 Measurement Process ................................................................................................223.3 Testing ........................................................................................................................233.4 Measuring Equipment .................................................................................................243.5 Measurement Deviations and Measurement Uncertainty ...........................................27

3.5.1 Bias ...............................................................................................................283.5.2 Repeatability .................................................................................................293.5.3 Reproducibility...............................................................................................303.5.4 Linearity.........................................................................................................313.5.5 Measurement Stability...................................................................................33

4 Influencing Factors on the Measurement Process ..................................................344.1 Typical Influencing Factors .........................................................................................344.2 Impact of the Influencing Factors ................................................................................374.3 Evaluation of the Measurement Process ....................................................................40

5 Gage Capability as a Measurement Process Capability Study ...............................445.1 Basic Procedures and Methods ..................................................................................445.2 Evaluation of Gages....................................................................................................47

5.2.1 Uncertainty of the Standard Master / Calibration Master ..............................475.2.2 Influence of the Resolution............................................................................495.2.3 Evaluation of the Bias ...................................................................................515.2.4 Study Type 1 .................................................................................................535.2.5 Quality Capability Indices Cg and Cgk ............................................................585.2.6 Study Type 1 for Characteristics with Unilateral Tolerances.........................665.2.7 Study Type 1 for Several Characteristics......................................................695.2.8 Linearity.........................................................................................................69

5.3 Evaluation of the Measurement Process ....................................................................80

x Table of Contents

5.3.1 Range Method (Short Method) .....................................................................805.3.2 Study Type 2: %R&R with Operator Influence..............................................825.3.3 Study Type 3: %R&R without Operator Influence.......................................102

5.4 Testing Measurement Stability .................................................................................1055.5 Further Studies .........................................................................................................109

5.5.1 Study Type 4...............................................................................................1095.5.2 Study Type 5...............................................................................................111

5.6 Method according to CNOMO ..................................................................................114

6 Capability Study of Attribute Measurement Processes..........................................1176.1 Attribute Gages.........................................................................................................1176.2 Attribute Gaging or Variable Measuring....................................................................1186.3 Requirements for Successful Inspections by Attribute .............................................1196.4 Analysis of Attribute Measurement Processes “Short Method” ................................1206.5 Analysis of Attribute Measurement Processes “Extended Method“..........................123

6.5.1 Introduction .................................................................................................1236.5.2 Testing Hypotheses ....................................................................................1276.5.3 Evaluating the Effectiveness of an Attribute Measurement System ...........1336.5.4 Signal Recognition Method.........................................................................137

7 Extended Measurement Uncertainty .....................................................................1437.1 Guide to the Expression of Uncertainty in Measurement .........................................143

7.1.1 Basic Principles ..........................................................................................1437.1.2 Aim and Purpose of the GUM.....................................................................1447.1.3 Field of Application .....................................................................................1457.1.4 Contents of the Guide.................................................................................1467.1.5 Terms and Definitions.................................................................................147

7.2 Determination of Measurement Uncertainties ..........................................................1507.2.1 Determination of the Standard Uncertainty.................................................1517.2.2 Determination of the Combined Standard Uncertainty ...............................1567.2.3 Determination of the Extended Uncertainty ................................................1587.2.4 Logging of the Uncertainty ..........................................................................1617.2.5 Expression of the Result .............................................................................162

7.3 GUM H.1 Example: Gage Block Calibration .............................................................1637.3.1 Measuring Task ..........................................................................................1637.3.2 Standard Uncertainties ...............................................................................164

7.4 Calibration of a Weight for the Nominal Value of 10 kg (S2) ....................................1727.4.1 Measuring Task ..........................................................................................1727.4.2 Standard Uncertainties ...............................................................................1727.4.3 Extended Measurement Uncertainty and Complete Measurement Result .179

7.5 Calibrating a Caliper .................................................................................................1817.5.1 Measuring Task ..........................................................................................1817.5.2 Standard Measurement Uncertainty (S10.3-S10.9)....................................1827.5.3 Extended Measurement Uncertainty and Complete Measurement Result .185

7.6 GUM Interpretation for Measurement Processes in Series Production ....................187

8 Extended Measurement Uncertainty according to ISO 22514-7 or VDA 5 ............1888.1 VDA 5 Flow Chart .....................................................................................................188

8.1.1 Schematic Approach...................................................................................1898.1.2 Gage Capability ..........................................................................................190

Table of Contents xi

8.1.3 Determination of the Standard Uncertainty as per Determination Method A1918.1.4 Determination of the Standard Uncertainty as per Determination Method B192

8.2 Principal Standard Uncertainty Components ............................................................1948.2.1 Standard Uncertainty uCAL ...........................................................................1968.2.2 Standard Uncertainty of the Resolution uRE ................................................1968.2.3 Standard Uncertainty uBI .............................................................................1978.2.4 Standard Uncertainty uMS in Case of Standard Gages................................1988.2.5 Standard Uncertainty Caused by Equipment Variation at the Reference Part

uEVR..............................................................................................................1998.2.6 Standard Uncertainty Caused by Equipment Variation at the Object uEVO..1998.2.7 Standard Uncertainty Caused by the Operator Influence uAV .....................2018.2.8 Standard Uncertainty Caused by the Test Object uOBJ ...............................2018.2.9 Standard Uncertainty Caused by the Temperature Influence uT.................2048.2.10 Standard Uncertainty Caused by Non-linearity uLIN.....................................2078.2.11 Standard Uncertainty Caused by Stability uSTAB..........................................208

8.3 Multiple Consideration of Uncertainty Components ..................................................2108.4 Determination of the Extended Measurement Uncertainty .......................................2118.5 Consideration of the Extended Measurement Uncertainty at the Specification Limits2118.6 VDA 5 Case Studies .................................................................................................213

8.6.1 Example: “Linear Measurement Using a Standard Gage” ..........................2138.6.2 Example: “Linear Measurement Using a Particular Gage”..........................220

9 Simplified Determination of the Measurement Uncertainty....................................2279.1 AIO Procedure (“All-in-One” Procedure) ...................................................................227

9.1.1 Measurement Process Capability Study .....................................................2279.1.2 Determination of the Extended Measurement Uncertainty..........................227

9.2 Practical Examples of the “All-in-One” Procedure.....................................................2319.2.1 Measurement Process with Linear Material Measure .................................2319.2.2 Measurement Processes without Linear Material Measure ........................233

10 Special Cases in Measurement Process Capability ..............................................23610.1 What Is a Special Case?...........................................................................................23610.2 Typical Special Cases...............................................................................................236

11 How to Handle Incapable Measurement Processes..............................................23811.1 Procedure for Improving Measurement Processes...................................................238

12 Typical Questions about Measurement Process Capability ..................................24112.1 Questions ..................................................................................................................24112.2 Answers ....................................................................................................................241

13 Capability Studies in Visual Inspections ................................................................24413.1 Requirements for Visual Inspections.........................................................................24413.2 Aptitude Test for Visual Inspectors ...........................................................................245

14 Purchase of Gages................................................................................................24814.1 Example for a Measuring Task Description ..............................................................24914.2 Example for a Requirement Specification .................................................................250

15 Proof of Suitability for Test Software .....................................................................25115.1 General Consideration ..............................................................................................25115.2 The Myth of “Excel Tables“ .......................................................................................254

xii Table of Contents

15.3 Gage Capability Test Examples ...............................................................................257

16 Appendix ...............................................................................................................27016.1 Tables ......................................................................................................................270

16.1.1 d2* Table for the Determination of k Factors and Degrees of Freedom fort Values.......................................................................................................270

16.1.2 Capability Limits according to VDA 5..........................................................27316.1.3 k Factors .....................................................................................................274

16.2 Analysis of Variance Models.....................................................................................27416.2.1 Measurement System Analysis – Study Type 2 .........................................27416.2.2 Measurement System Analysis – Study Type 3 .........................................279

17 Reference..............................................................................................................28217.1 Abbreviations ............................................................................................................28217.2 Formulas...................................................................................................................28617.3 Bibliography ..............................................................................................................28817.4 Figures......................................................................................................................29617.5 Tables ......................................................................................................................300

18 “Measurement System Capability” Reference Manual ..........................................30119 GM PowerTrain Measurement Systems Specification (SP-Q-MSS)......................32920 Bosch Booklet 10: Capability of Measurement and Test Processes......................40221 Index......................................................................................................................428

1

1 Measurement Process Capability

1.1 Introduction

1.1.1 Why Measurement Process Capability?

One sentence can answer this question. “It is the necessity to have appropriatemeasurement processes available for the correct evaluation of manufacturing andproduction processes.” The measurement values determined in the measurementprocess are the basis of the evaluation and have to reflect the actual situation quiterealistically. A non-capable measurement process does not give a true picture of realityand does not allow reliable conclusions to be drawn.

Hence, the following question has to be answered: “What does a capable measurementprocess mean?” Today, several standards and guidelines (see Figure 1-1), answer thisquestion. These standards and guidelines require capability studies and describeprocedures how to conduct them.

StandardizationRequirements

AutomotiveRequirements

DIN VEN 13005 (GUM)

DIN EN ISO 14253

DIN 1319

DIN 2257

DIN EN ISO 9000ff

DIN ISO 10012

ISO/TS16949

VDA 6.1

VDA 5

VDI / VDE

DGQ / DKD

Company Guidel ines

QS-9000

MSA

Q-DASGuideline

UUy

ToleranceLSL USL

UUy

ToleranceLSL USL

ToleranceLSL USL

UUy

Figure 1-1: Important requirements of automotive and standardizationguidelines in connection with measurement process capability

Figure 1-1 displays important requirements of automotive and standardizationguidelines.

2 1 Measurement Process Capability

For the evaluation of measurement systems, ISO/TS 16949:2002 [37] requires:

“Each kind of measurement system demands statistical studies in order to analyze thevariation of the measurement results. This requirement applies to all measurementsystems that are referred to in the production control plan. The applied methods andacceptance criteria have to correspond to the criteria mentioned in the customer’sreference guide for the evaluation of measurement systems. Other analytical methodsand acceptance criteria may only be applied on the customer’s authority.”

The statement that other methods are possible on the customer’s authority is notrelevant to many suppliers because they cannot make individual agreements with alltheir customers. In order to certify the QM system, they can only consult generalstandards (e.g. MSA [1] or VDA [70]) and use them as the basis for process capability.

Chapter 1.2 “Historical Retrospect and Prospect“ illustrates the context and thedevelopment of the indivdual documents. Machine and production facilities acceptance,evaluation of processes and products or continuous process monitoring are based onthe evaluation of qualitative and quantitative product characteristics. The analysisfocuses on quantitative or variable characteristics. However, another chapter deals withcapability studies for qualitative or attribute characteristics.

In case of qualitative characteristics, measurement systems provide measurementvalues of the characteristics of the produced parts or process parameters. This requirestask-related measurement systems, specific sensors and customary standardmeasuring devices.

In order to draw the right conclusions from the measurement values, the values must berecorded with sufficient measurement accuracy“ relating to the characteristic’s toleranceor the process. In the past, the suitability of a measuring device was tested by means ofminimum values given in standards or the manufacturer’s specifications weremonitored. Today, clear specifications are available. The ISO 10012:2004 [29] requiresthe determination of the measurement uncertainty according to EN 13005 [32].However, testing the gage under ideal conditions is only one single component whendetermining the measurement uncertainty of the measurement process, e.g. in themetrology lab, tested by trained staff, using idealized parts such as a standard masteror calibration master, and in standardized facilities. The VDI / VDE / DGQ guideline2618 [74] provides examples of the procedure and test methods in the form of testinstructions. Continuous monitoring (gage monitoring) and new devices for theinspection of manufacturer’s specifications require this proceeding in order to detectchanges or errors at the device.

The determined “capability“ does not reveal anything or hardly anything about thebehavior of the device under real conditions (see Figure 1-2).

1.1 Introduction 3

Measurement resultand measurementdeviation

Man

Environment

Part

Gage

Method

MeasuringstrategyMeasurementprocedure

Measurementprinciple

Motivation

EvaluationEfficiency

Humidity ResolutionVibrations Device specific evaluation

TemperatureStructure

Statistical behavior

Oscillations Dynamic behavior

Surface conditioni.e. roughness

Form deviations

Mass

Dimensions

Figure 1-2: Influencing factors on a measurement resultSource: Pfeifer, T.: Fertigungsmesstechnik, R. Oldenbourg

NoteRelevant experience has shown that the influence of a measuring device on themeasurement result is only a minor component in the entire measurement process.When examining devices, all influencing factors must be taken into account.

At best, the procedure described above can give a theoretical statement about whetheror not a measuring device might be suitable for a specified tolerance. In order to find outwhether or not the measurement system is capable or qualified to make a realisticevaluation of a process including very little variation and considering the above-namedinfluencing factors, other procedures and methods are required.

The aim of “never-ending improvement” leads to the fact that tolerances becomenarrower and the process variation becomes smaller. For this reason, the measurementsystem must fulfill the respective task. If this is not the case, the measurement resultswill be wrong and they will become almost useless for statistical analysis.

In order to meet these requirements, several standards and guidelines demand thedetermination of the measurement uncertainty or the evaluation of the measurementsystems by means of capability studies. ISO/TS 16949 [37] also obliges processcapability studies but it does not specify a certain procedure. It only says that thecustomer specification must be applied, either gage capability studies or thedetermination of the extended measurement uncertainty. In the near future, bothprocedures will most likely be on the same par. In case of the American automotiveindustry, MSA will still be used. As regards the German automotive industry, thedetermination of process capability according to VDA 5 is due.

4 1 Measurement Process Capability

This book deals with the different procedures and approaches. Moreover, it discussesthe connections and differences between these procedures.

Figure 1-3 and Figure 1-4 show how the “variation of the measurement system” affectsthe “observed process variation.” In Figure 1-3, the “variation of the measurementsystem” is quite small. In this case, the “actual process variation” is almost identical tothe “observed process variation.” The “variation of the measurement system” displayedin Figure 1-4 is too large. There is a clear difference between the “actual processvariation” and the “observed process variation.” Figure 1-4 shows only one direction ofthis difference (upwards) for the sake of clarity. However, the “observed processvariation” can also deviate downwards. This difference leads to the misinterpretation ofthe actual situation.

observedprocess variation

small variationmeasurement system

actualprocess variation

Figure 1-3: Observed process variation showing small variation of themeasurement system

observedprocess variation

large variationmeasurement system

actualprocess variation

Figure 1-4: Observed process variation showing the variation of themeasurement system

Now, there are two questions:1. How can the “variation of the measurement system” be determined?2. What is the maximum value of the “variation of the measurement system” in

order to gain an acceptable difference between the “observed process variation”and the “actual process variation”?

1.1 Introduction 5

The first question can be answered by the procedures described in the next sections.Depending on the procedure, various statistical values are calculated, such as Cg,%R&R, U etc. The problem described in the second question can be solved bycomparing the results to specified limit values. Figure 1-5, Figure 1-6 and Figure 1-7show how the statistical values affect the process capability Cp.

U/T0.50.40.30.20.10

MeasuredProcessCapabilityCp

0

0.5

1.5

2.5

1

2

Measurement Process Capability RangeType-1 Study: C = 1.0 ... 1.67)p

Golden RuleRange

Measurement ProcessCapability RangeType 2/3 Study)

Cp* = 2

Cp* = 1.33

Figure 1-5: Influence of U/T on quality capability index Cp

0

1

2

3

4

5

6

0,5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0

Observed C Valuep

RealC

Value

p

70 % 60 % 50 %

40 %

30 %20 %10 %

%R&R

Figure 1-6: Influence of %R&R on quality capability index Cp

with tolerance as reference figure

6 1 Measurement Process Capability

Cp is a quality capability index as defined in [13]. The approach can be translated intoother quality capability indices, such as Cm, Pp or Tp.

Figure 1-5 shows how the measurement uncertainty affects the quality capability indexCp. It displays the run chart of the Cp value (for Cp = 2.0 and Cp = 1.33) using the ratiobetween the measurement uncertainty and the tolerance (U/T). In case of U/T = 0.15(i.e. U amounts to 15% of the tolerance), the actual Cp value drops from 2.0 to 1.5!

0 0.5 1 1.5 2 2.5 3Observed C Valuep

0

0.5

1

1.5

2

2.5

3

RealC

Value

p

10 %

30 %

50 %

70 %

90 %

%R&R

Figure 1-7: Influence of %R&R on quality capability index Cp withprocess variation as reference figure

Figure 1-6 and Figure 1-7 display the change of the Cp value plotted against thestatistical value %R&R. In both cases, the difference between the “actual Cp value” andthe “observed Cp value” amounts to

up to %R&R = 10% very small up to %R&R = 20% or 30% barely acceptable

When the %R&R value amounts to more than 30%, the deviation of the “observed Cpvalue” from the “actual Cp value” is too sharp and cannot be accepted. For this reason,most guidelines demand that the measurement system is only “capable” up to %R&R =10%. If the statistical value %R&R lies within the range of 10% to 30%, it is referred toas “conditionally capable” and must not be used for the measuring task.

If a measurement process is not capable, the recorded measurement values for theevaluation of machine capability and preliminary and ongoing process capability will beuseless. This method was already introduced in the middle of the 1980s, however,without observing the capability of measurement processes. The variation of themeasurement process was normally evaluated and not the manufacturing process itself.This was the main reason for the introduction of gage capability.

1.2 Historical Retrospect and Prospect 7

“You can only produce as accurately as you can measure” is a common statement.Many practitioners claim that this is wrong. The reason for this opinion is that themeasurement procedures are not sufficiently accurate. However, the measurementprocedure requirements will further increase. The specification limits often becomenarrower for technical reasons. Hence, the production-related margin becomes smallerwhich makes reasonable rules and controls important. These are based on themeasurement results of a measurement process. A typical example of highest demandson a measurement process is the production of Common Rail fuel injectors. In order toreach pressures of about 1600 bar, the roundness of pins and bores of the order of lessthan 0.5m is required. Such a production can only be managed at a affordable costrate when the characteristics are controlled by a suitable measurement process. Thisexample can be translated into many other applications.

The feasibility of the measurement procedure must be already confirmed whenengineering a process. This is quite important for the determination of specificationlimits. A team consisting of engineering staff, manufacturing staff and customers orsuppliers shall determine these limits. Only in this way can aims be achievedeconomically.

1.2 Historical Retrospect and Prospect

The fact that a measurement result shows a certain uncertainty is as old asmeasurements themselves. In practice, the slogan “Who measures, ..measuresrubbish” is often used. This simply means that the measurement uncertainty is too largewhich leads to a wrong result. In the past, the measurement uncertainty often was alsoreferred to as error.

The German Committee for Measurements and Weight was the first to recommend amore detailed definition of “measurement uncertainty”. This recommendation was givenin 1977. In 1994, the term was specified and defined in the first edition of “InternationalVocabulary of Basic and General Terms in Metrology” (VIM) [49]. The definition ofmeasurement uncertainty as it is applicable today can also be found in DIN 1319-Part 1[20]. In 1995, the “Guide to the Expression of Uncertainty in Measurement” (GUM)explained exactly how to determine the measurement uncertainty. This guide has beenused as pre-standard DIN V EN V 13005 [32] since 1999.

Although people have known for a long time that it is necessary to determine theuncertainty of the respective measurement process, many documents were buried inthe files. The reason why is easy to explain. Taking a look at DIN V EN V 13005 [32] thedocument includes many theoretical considerations that are hard to realize in practice.Metrology labs have dealt with this subject for a long time because of the requiredcertification or accreditation. However, the problem is easier to tackle there because ofthe limited number of influencing factors. Measurement processes including manyinfluencing factors, as they exist in everyday practice, cannot be analyzed andevaluated just like that taking a theoretical approach. The individual influencing factors