MGPG 41

CMM Measurement
Strategies
David Flack
No. 41
The National Physical Laboratory is operated on behalf of the DTI by NPL Management Limited, a wholly owned subsidiary of Serco Group plc

Measurement Good Practice Guide No. 41

CMM Measurement Strategies

David Flack
Dimensional and Optical Metrology Team
Centre for Basic, Thermal and Length Metrology

Abstract: This guide covers the selection of the number of measurement points when
using Co-ordinate Measuring Machines (CMMs) and gives advice on the compromise
between accuracy and speed. It provides guidance on sampling criteria for standard
features and advice on measurements that involves projections of features over long
distances. It covers cleanliness, part loading/alignment and the effect of temperature,
surface finish and geometry on the final result. It also contains information on basic
measurement principles, common measurement requirements, CMM software
functionality in relation to drawing requirements and good metrology practice when
using CMMs with CAD data to inspect parts.

Crown Copyright 2001
Reproduced by permission of the Controller of HMSO

ISSN 1368-6550

July 2001

National Physical Laboratory
Teddington, Middlesex, United Kingdom, TW11 0LW

Acknowledgements

The author acknowledges the following contributions to this guide. Alan Hatcher and
Keith Bevan of Mitutoyo for providing useful technical input. Professor Maurice Cox of
NPL for the sections on the partial arcs and the scientific approach to probing strategies.
Dr. Graham Peggs and Dr. Robert Angus for many suggested changes to the manuscript.
D. R. Coleman and T. F. Walters for permission to use images from their book
Fundamentals of Touch Trigger Probing. UKAS for their valuable input to the project. All
lead users who reviewed early drafts of this guide and last but not least the Department of
Trade and Industry (DTI) for funding production of this guide as part of the 1999-2002
Length Program (Project MPU 8/61.3).

CMM Measurement Strategies

Contents
1. INTRODUCTION.........................................................................................................1
1.1 CO-ORDINATE MEASURING MACHINES..............................................1
2. GENERAL MEASUREMENT STRATEGY .............................................................3
3. SELECTION OF THE FEATURES TO BE MEASURED.......................................4
3.1 FEATURES........................................................................................................4
3.2 SETUP................................................................................................................4
3.3 REPOSITIONING METHODS.......................................................................5
3.4 SUMMARY.......................................................................................................5
4. DEFINE THE WORKPIECE DATUM TO BE USED WITHIN THE CO-
ORDINATE SYSTEM............................................................................................................6
4.1 INTRODUCTION............................................................................................6
4.2 DATUM FEATURES ON DRAWING..........................................................6
4.3 CHANGING THE DATUM.........................................................................10
4.3.1 Introduction...........................................................................................10
4.3.2 Procedure ..............................................................................................10
4.3.3 An example...........................................................................................10
4.4 SMALL FEATURES.......................................................................................15
4.5 SUMMARY.....................................................................................................16
5. SELECTION OF THE WORKPIECE ORIENTATION........................................16
6. SELECTION OF THE WORKPIECE HOLDING METHOD.............................17
6.1 MEASUREMENT FORCE............................................................................17
6.2 HEAVY ITEMS...............................................................................................17
6.3 LIGHT ITEMS.................................................................................................17
6.4 FIXTURES .......................................................................................................18
6.5 ADHESIVES ...................................................................................................18
6.6 GENERAL.......................................................................................................19
6.7 SUMMARY.....................................................................................................19
7. PROBE QUALIFICATION.......................................................................................20
7.1 BASIC QUALIFICATION............................................................................20
7.2 ADDITIONAL CHECKS ..............................................................................22
7.3 STYLUS CHOICE ..........................................................................................23
7.4 COMPARISON (SUBSTITUTION METHOD) ..........................................24

7.5 SUMMARY.....................................................................................................24
8. DEFINITION OF THE PROBING STRATEGY ...................................................25
8.1 INTRODUCTION..........................................................................................25
8.2 DISTRIBUTION OF POINTS .......................................................................27
8.2.1 The AD HOC approach.....................................................................28
8.2.2 The scientific approach.........................................................................39
8.3 PARTIAL FEATURES AND PARTIAL ARCS ..........................................42
8.4 SUMMARY.....................................................................................................44
9. PROBING CONSIDERATIONS IN THE MEASUREMENT STRATEGY.....45
10. PROGRAMMING THE CMM.................................................................................45
10.1 USER REQUIREMENT.................................................................................45
10.2 FUNCTIONAL SPECIFICATION...............................................................45
10.3 TESTING.........................................................................................................46
10.4 GENERAL.......................................................................................................46
10.5 SUMMARY.....................................................................................................47
11. CMM DATA ASSESSMENT SOFTWARE ...........................................................48
12. SOFTWARE FUNCTIONALITY .............................................................................50
12.1 NON-CONVENTIAL PART ALIGNMENT..............................................51
12.2 CAD INTERFACE .........................................................................................51
12.3 GRAPHICAL FEATURE ANALYSIS .........................................................51
12.4 AUTOMATIC FEATURE MEASUREMENT.............................................51
12.5 REVERSE ENGINEERING...........................................................................52
12.6 OFFLINE PROGRAMMING........................................................................52
12.7 PROGRAMMING LANGUAGE.................................................................52
12.8 UNKNOWN GEOMETRY SCANNING....................................................52
12.9 GEOMETRIC TOLERANCING...................................................................53
12.10 COMPLEX SURFACE MEASURING.........................................................53
12.11 INTEGRATED BEST FIT ..............................................................................53
12.12 INTEGRATED STATISTICS.........................................................................53
13. ASSESSMENT INFORMATION RECORD..........................................................55
14. ALIGNMENT CRITERIA FOR STANDARD FEATURES ................................56
14.1 LENGTH.........................................................................................................56
14.2 DIAMETER AND POSITION OF A HOLE...............................................57
14.3 DIAMETER AND POSITION OF A SHAFT .............................................59
14.4 AN ANGLE BETWEEN TWO LINES.........................................................60
15. ENVIRONMENTAL CONDITIONS......................................................................62

15.1 TEMPERATURE ............................................................................................62
15.2 VIBRATION....................................................................................................63
15.3 DUST................................................................................................................64
15.4 SUMMARY.....................................................................................................64
16. SUMMARY..................................................................................................................65
17. GLOSSARY OF TERMS............................................................................................66
18. HEALTH AND SAFETY............................................................................................68

APPENDIX A LINKS TO OTHER USEFUL SOURCES OF INFORMATION.69
A.1 NATIONAL AND INTERNATIONAL ORGANISATIONS...................69
A.1.1 National Physical Laboratory...............................................................69
A.1.2 National Institute of Science and Technology (NIST).........................69
A.1.3 EUROMET, A European Collaboration in Measurement Standards ..69
A.1.4 European Co-operation for Accreditation (EA)...................................70
A.2 CLUBS .............................................................................................................71
A.2.1 Dimensional Metrology Awareness Club (DMAC) ............................71
A.2.2 Software Support for Metrology Programme (SSfM) .........................71
A.3 NATIONAL AND INTERNATIONAL STANDARDS............................72
A.3.1 British Standards Institution (BSI).......................................................72
A.3.2 International Organisation for Standardisation (ISO) ..........................72
A.4 TRACEABILITY.............................................................................................73
A.5 NATIONAL MEASUREMENT PARTNERSHIP......................................74
A.6 TRAINING COURSES ..................................................................................75
A.7 AWARDING BODIES...................................................................................75
A.8 FURTHER READING...................................................................................76
APPENDIX B USER REQUIREMENT..........................................................................77
APPENDIX C FUNCTIONAL SPECIFICATION.......................................................79
APPENDIX D TESTING..................................................................................................84

List of Figures

Figure 1 Multiple set ups........................................................................................................5
Figure 2 Finding the centre line.............................................................................................7
Figure 3 Cartesian co-ordinate system ( D. R. Coleman & T. F. Waters) .....................8
Figure 4 The six touch points needed to create a coordinate system ( D. R. Coleman
& T. F. Waters)................................................................................................................9
Figure 5 The component functional features.....................................................................11

Figure 6 The functional features..........................................................................................12
Figure 7 New datum.............................................................................................................13
Figure 8 Comparison of limits.............................................................................................14
Figure 9 The new dimensions..............................................................................................14
Figure 10 Initial tolerances...................................................................................................15
Figure 11 Final tolerances.....................................................................................................15
Figure 12 Effect on incorrect alignment of cylinder end face .........................................15
Figure 13 The probe configuration on a datum sphere for stylus tip calibration
(courtesy F M M Chan et al) .......................................................................................21
Figure 14 Probe contacts on gauge blocks and ring gauge (courtesy F M M Chan et
al) ....................................................................................................................................22
Figure 15 (Effective Working Length source Renishaw Technical Specification
Manual) .........................................................................................................................23
Figure 16 Defining common geometric features (courtesy Mitutoyo) ..........................26
Figure 17 A distribution of points on a line.......................................................................28
Figure 18 A distribution of points in a plane ....................................................................30
Figure 19 A chess board distribution of points in a plane ............................................30
Figure 20 six uniformly spaced points (*) with complete failure to detect lobing.......31
Figure 21 Seven uniformly spaced points (*) with at least 79% of the lobing detected
........................................................................................................................................32
Figure 22 A distribution of points on a sphere .................................................................33
Figure 23 A distribution of points on a cylinder...............................................................34
Figure 24 Hole not square to datum plane XY ( D. R. Coleman & T. F. Waters) ......36
Figure 25 A distribution of points on a cone.....................................................................38
Figure 26 Ten uniformly spaced measurements on a partial arc, showing the
(nominal) circle of which it is part.............................................................................43
Figure 27 Average size (Gaussian) and enveloping sizes (Chebyshev)( D. R.
Coleman & T. F. Waters).............................................................................................49
Figure 28 Two probing points for measuring a length ( D. R. Coleman & T. F.
Waters)...........................................................................................................................56
Figure 29 Cylindrical polar coordinate system ( D. R. Coleman & T. F. Waters) .....57
Figure 30 Three probing points for measuring hole diameter ( D. R. Coleman & T.
F. Waters) ......................................................................................................................58
Figure 31 Three points for measuring shaft diameter ( D. R. Coleman & T. F.
Waters)...........................................................................................................................60
Figure 32 Four probing points for simple angle measurement ( D. R. Coleman & T.
F. Waters) ......................................................................................................................61

List of tables

Table 1 number of contact points required........................................................................26

MEASUREMENT GOOD PRACTICE

There are six guiding principles to good measurement practice that have been defined by
NPL. They are

The Right Measurements: Measurements should only be made to satisfy agreed and well-
specified requirements.

The Right Tools: Measurements should be made using equipment and methods that have been
demonstrated to be fit for purpose.

The Right People: Measurement staff should be competent, properly qualified and well informed.

Regular review: There should be both internal and independent assessment of the technical
performance of all measurement facilities and procedures.

Demonstratable Consistency: Measurements made in one location should be consistent with
those made elsewhere

The Right Procedures: Well-defined procedures consistent with national or international
standards should be in place for all measurements


1

1. INTRODUCTION
1.1 CO-ORDINATE MEASURING MACHINES

Co-ordinate metrology has become essential for industrial dimensional metrology. Goods
worth over 100M are inspected annually, in the UK alone by thousands of co-ordinate
measuring machines. It is therefore essential that the accuracy can be estimated and
traceability demonstrated. In addition quality systems have brought about increased user
awareness of the benefits provided by reliable and frequent checks of CMMs. Over the
years, standards and guidelines have been developed to harmonize the performance
specifications of a CMM to enable a user to make meaningful performance comparisons
when purchasing a machine and, once purchased, to provide a well-defined way in which
the specified performance can be reverified. For the user, demonstrating traceability to
national standards and estimating the accuracy of measurements made with three
dimensional CMMs is of extreme importance for maintaining confidence and reliability in
the measurements.

The ISO 10360 series of standards detail the acceptance, reverification tests and interim
checks required to determine whether the CMM performs to the manufacturers stated
error of indication. However even with these tests it is not possible to make a statement
about the length measurement capability of the machine due to the complicated way in
which the uncertainties associated with the CMM combine. Therefore, the length
measurement uncertainty derived from a limited sample of measurements cannot be
considered to be representative of all the possible length measurement tasks and certainly
not of the measurement tasks the CMM is capable of performing. In effect the tests do not
guarantee traceability of measurement for all measurement tasks performed. The user
should be aware of this important fact and develop task-related measuring strategies for
each measurement undertaken that will provide the appropriate level of confidence in the
overall result. Virtual CMMs can help meet this requirement. Further information on
virtual CMMs can be found in NPL report CMSC 01/00 Simulated Instruments and
Uncertainty Estimation A B Forbes and P M Harris. CMM verification is the subject of a
separate best practice guide.

The sampling strategy for a CMM inspection process is under the users control whereas
the accuracy level is associated with the machine and the software. Over the last few years
the advancement of hardware and software technology has resulted in machines that are
potentially capable of high accuracy measurements. However, the inspection quality and
hence the confidence in the result can be impaired by improper measurement strategies.


2
The selection of a measuring strategy that creates confidence in the final result demands
skill, experience and attention to detail on the part of the user. This guide sets out to assist
the user in identifying the uncertainties inherent in the use of CMMs and advising on the
strategies that will provide confidence in the measured results.

In contrast with simple, single-purpose measuring instruments, CMMs are able to
measure a wide range of geometrical parameters. For each of these parameters the user
may adopt any of a number of measurement strategies. These include the selection of a
particular probe and stylus configuration, the number and position of measuring points
and the direction and speed of approach of the probe. Both the measurement task and the
selected measurement strategy determine the way in which errors are introduced in to the
measurement system to influence the uncertainty associated with the measurements.


3

2. GENERAL MEASUREMENT STRATEGY

The general measurement strategy can be subdivided into a number of parts that should
be followed by the user:

1. Selection of the features on the workpiece to be measured.

2. Definition the workpiece datum(s) to be used within the co-ordinate system.

3. Selection of the workpiece orientation.

4. Selection of the workpiece holding method.

5. Qualification of the probe

6. Definition of the probing strategy.

7. Programming of the CMM.

8. Assessment information record

The following sections will cover each of these points in turn.

4

3. SELECTION OF THE FEATURES TO BE MEASURED.
3.1 FEATURES

In general, production and functionality requirements will determine which features need
to be measured. In some cases a component could have some features that (a) cannot be
measured on a CMM due to accessibility problems (b) are impractical to measure such as
circular sections of small arc length and (c) could be more cost effectively measured using
other instruments. Consideration should therefore be given to the minimum number of
surfaces that require CMM measurement in order for the user to establish the accuracy of
the workpiece.

3.2 SETUP

In general it is important to establish a strategy that requires, if possible, a single set up of
the workpiece for the measurement of all relevant features. If more than one set up is used
there is the possibility of operator error degrading the measurement process and hence
reducing the confidence in the final results. However, on the other hand, using more than
one set up can enhance the accuracy by minimising the need for complicated stylus
arrangements. In general the user should use the minimum number of set ups for the
measurement of a workpiece. One set up should be the aim. The way in which any
measurement is made is generally a compromise between speed and accuracy. To achieve
the accuracy required for some tight tolerances may require multiple set-ups and hence
longer measurement time. Always bear in mind the target tolerance. This is illustrated in
the following example.

In Figure 1 the distances between the spheres A, B and C are required. These distances can
be measured in one set up as shown. The distance BC will probably have a smaller
uncertainty than the other two lengths as only one machine axis is used for this
measurement. Measurement of length BA and AC require the use of two machine axes
hence the likelihood of a larger uncertainty. However if the item is rotated so that the
other lengths fall in place of length BC, the three set-ups will result in the lengths being
known with smaller uncertainty because the same scale and location are used for each
measurement.

5

Figure 1 Multiple set ups

3.3 REPOSITIONING METHODS

A further extension to this idea in section 3.2 is the technique known as re-positioning.
With this technique reference artefacts such as spheres are attached to the item being
measured. The item can then be measured using multiple set-ups. For each set up the
positions of the reference artefacts are also measured. Software is then used to collate all
the data from the various set-ups into the same frame of reference (co-ordinate system).
Repositioning is described in further detail in NPL report CLM2 Measurement of Artefacts
using Repositioning Methods.

3.4 SUMMARY

Use minimum number of set-ups
Use only one axis for critical dimensions if possible
Choose the most appropriate machine for the job
A
B C

6

4. DEFINE THE WORKPIECE DATUM TO BE USED WITHIN
THE CO-ORDINATE SYSTEM.

4.1 INTRODUCTION

A datum is a reference surface, plane, line or point used to facilitate the definition of
features on a part. A datum is usually a locating or positioning feature used for (non-
CMM) measurement. It is also the origin from which the location or geometric
characteristics of a part are established. A datum feature is an actual feature of a part that
is used to establish a datum and can be a surface, an axis or a median plane. A functional
datum feature is a surface on a component, which is of importance to the function of the
component in an assembly. Definition of a datum can be a complicated subject and the
reader is referred to standards ASME Y14.5M, ISO 1101 and ISO 5469. Understanding the
datum is important as large errors can result from using an inappropriate datum.

In the preparation for measurement, using a CMM, of a component based on a detail
drawing derived from the assembly drawing, it is necessary to identify functional datum
features - those features that are critical in the function of the component. These features
will have small tolerance dimensions. A component detail drawing must present
information in a form that is suitable for manufacture (and inspection), and must show the
required manufacturing dimensions that are related to the datum(s).

4.2 DATUM FEATURES ON DRAWING

In most cases the co-ordinate system should make use of datum features identified in
drawings or technical documents relating directly to the workpiece.

However, a workpiece datum can be developed from theoretical points that cannot be
measured by the CMM; these are usually points indicated on mating part surfaces or co-
ordinates used in a large assembly.

It is not usually the prerogative of the designer to decide the details of the machining of a
component, although it is often possible to foretell the sequence of some of the
manufacturing processes involved. From knowing this sequence the designer can identify
the manufacturing datum face(s), and from this, the required machining dimensions.


7
A datum feature is an important feature - a locating or positioning feature. A datum
feature can be a face (a surface) or the centre line of a hole. A functional datum feature is a
face, or a hole, in a component, which is of importance to the function of the component.

For example a cylinder centre-line cannot be measured, as it does not exist as a physical
entity, in this case the user would generally measure a functional aspect on the workpiece
such as a face or hole diameter and relate this position to the intended position of the
centre line.

In Figure 2 to determine the dimension (x) the user would first set the relevant co-ordinate
system and using the datum face would measure the diameter of the hole to establish the
theoretical centre-line of the hole.

Figure 2 Finding the centre line

When creating a datum it is preferable to choose the surfaces that are used in the
manufacturing process to hold the workpiece as datum features. This choice relates the
inspection results directly to the manufacturing process.

X
Datum
Face

8
The features of any workpiece can be defined in two ways, relative to a datum position or
positions, or relative to one another. The co-ordinate system should be clearly defined on
the engineering drawing.

The datum features are used to define a co-ordinate system. The most common co-
ordinate system in use is the Cartesian system (Figure 3). It comprises three linear axes
that intersect at the system origin. The three axes are mutually perpendicular and are
usually designated X, Y and Z.

Figure 3 Cartesian co-ordinate system ( D. R. Coleman & T. F. Waters)

Lord Kelvins principle of kinematics states that a body free to move in space has six
degrees of freedom. There are three degrees of freedom linearly along the X, Y and Z-axes
and three degrees of freedom in a rotational sense around the X, Y and Z-axes. Therefore
to define a co-ordinate system a minimum of six measurement points are required (Figure
4).

To define a datum system for a rectangular part with flat surfaces that has its origin in one
corner would require six contact points (Figure 4). Three points are required to define a
flat surface or plane (the XY plane primary datum A), a further two points are required
to define a line which lies on the XZ plane (secondary datum B) and a final point in the YZ
plane (tertiary datum C).


9

Figure 4 The six touch points needed to create a coordinate system ( D. R. Coleman & T. F. Waters)
Note that the three planes are created to be mutually perpendicular and the point of
intersection of these three planes is the origin of the system X = 0, Y = 0 and Z = 0.

The choice of datum(s) for a component and their precedence should normally be
specified on the product drawing. If the user decides on an alternative datum, the effect on
the results can be quite dramatic.

When considering the features that are required for the creation of the datum on the
component, the user should determine whether the primary datum is a surface or an axis.
The next stage is to determine the secondary and tertiary datum and from these the co-
ordinate system to be used for the measurement strategy can be set. A document on
datum systems is available form NIST A conceptual Data Model of Datum Systems Journal of
Research of the National Institute of Standards and Technology Volume 104, Number 4, July-
August 1999. It can be downloaded free of charge from the NIST web site.


10
4.3 CHANGING THE DATUM

4.3.1 Introduction

It should not be necessary for anyone using a drawing or technical documentation to have
to interpret the presented information, the user should not have to carry out any
calculations or adjustments in order to complete the activities relating to the component
measurement. However in some cases it may be necessary for the user of a CMM to relate
certain features of the workpiece to a datum not indicated in the drawing or technical
documentation. This can be a complex procedure, which will require a number of steps to
complete and in the course of which toleranced dimensions may have to be altered. Care
should be taken when making such alterations.

4.3.2 Procedure

The procedure to be followed when changing the datum face is a complex one and can be
summarised as:

1. Decide on the required datum face.
2. Decide on the required measuring dimensions (which may also involve deciding that some
dimensions will have to be omitted).
3. Determine the tolerance for each of the new dimensions to be measured.
4. Set suitable limits for all but one of the required dimensions.
5. Determine the limits of the final dimension.

4.3.3 An example

The example given here relates to changing a datum for a machining operation but this
example equally applies to measurement. In the preparation of a component detail
drawing, it is necessary to identify functional datum faces and functional holes, these faces
and holes will be critical in the functioning of the component. These features are
recognisable, as they will have small tolerance dimensions. In Figure 5 there are four
functional features.


11

Figure 5 The component functional features

The functional features are:

1. The 39.975/39.936 mm limits indicate that this diameter is to be a running fit in a
mating component.

2. The 15.027/15.000 mm limits indicate an important internal feature.

3. The two linear dimensions 35.3/35.0 mm and 15.00/14.85 mm are functional
dimensions in that they indicate that the bottom face of the groove is functionally
important, both to the right-hand end face of the component and the right-hand
face of the 60 mm diameter, therefore the bottom of the groove is a functional
datum face.


12

Figure 6 The functional features
The manufacturing sequence will commence with a turning operation followed by a
grooving operation. Therefore, for the turning operation a direct dimension, with
appropriate limits, is required for the length of the of the 39 mm diameter. These limits
cannot be obtained by adding the two chain dimensions together (35.3/35.0 mm and
15.00/14.85 mm).

The manufacturing datum face for the turning operation has to be the right-hand end face
of the component. This now will require a changing of the datum face, which as stated
above was the functional datum face at the bottom of the groove.

When the datum face is changed there is, inevitably, a reduction in tolerance of one, or
more, dimensions. Also, dimensions must never be duplicated. This means that in
showing the new dimension for the 39 mm diameter one of the chain dimensions
35.3/35.0 mm and 15.00/14.48 mm must be omitted from the drawing.

In the interests of obtaining the largest tolerance possible for the dimensions involved, the
dimension having the largest tolerance should be omitted i.e. the 35.3/35.0 mm
dimension, because the tolerance of the omitted dimension (35.3/35.0 mm) i.e. 0.3 mm is
the total tolerance for the remaining, controlling, dimensions (50.15/50.00 and
15.00/14.85).


13

Figure 7 New datum
In summary:

1. The manufacturing datum face, for both processes, is the right hand end face.

2. Dimension L for turning, dimension G for groove depth, the 35.3/35.0 mm
dimension omitted.

3. Tolerance of the omitted dimension is 0.3 mm. Let tolerance (limits) for G remain as
0.15 mm. Therefore, tolerance for L is 0.15 mm

4. Limits for G remain unaltered i.e. 15.00/14.85 mm.

5. Determine limits for L as follows: to ensure that the limits of omitted dimension O
are not exceeded (Figure 7), use is made of a diagrammatic representation of the
limits L and G in terms of the limits of O.


14

Figure 8 Comparison of limits
When O is minimum, then L is minimum and G is a maximum.
When O is maximum, then L is maximum and G is a minimum.

Hence:
L min = 35.0 + 15.0 = 50.0 mm
L max = 35.3 + 14.85 = 50.15 mm

Verification that the limits of dimension O, although now omitted from the drawing, will
not be exceeded, is shown in Figure 8.

The component will now require re-dimensioning to suit the requirements of the
operation sequence.

Figure 9 The new dimensions

15

Figure 10 Initial tolerances

Figure 11 Final tolerances

The tolerance reduction that occurs when the datum face is changed is shown in the new
dimension for the turning operation, where the tolerance of 0.15 mm is half the tolerance
of the omitted dimension.

4.4 SMALL FEATURES

Care should be taken when alignment is made on small features. Small features should not
be used as the primary datum. An example of this is aligning the end face of a long thin
cylinder. Any small error in aligning the end face (perhaps due to the face geometry) will
produce a progressively greater error when checking the axis of the cylinder, i.e., the
longer the cylinder the greater the error.

Effect of incorrect alignment of cylinder end face
Progres sive
erro r
Probing

Figure 12 Effect on incorrect alignment of cylinder end face


16
In cases where alignment is required on small features the user should assess the
functionality of the workpiece and align a chosen surface or datum in an appropriate
manner to avoid unnecessary geometric errors. In the case of the cylinder end face it may
be more appropriate to align the outer diameter of the cylinder and probe the end faces
from this aligned datum. It should be noted that the use of a non-specified datum might
cause errors in the final results. A non-specified datum is one that has not been specified
by the designer and which the user has, in the interests of accuracy or ease of
measurement, substituted a functional datum feature as new datum. Since in some cases
the use of an alternative datum may be unacceptable the user should discuss the
implications of changing a datum with the design department.

4.5 SUMMARY

Use datum on drawing if possible
Take care when changing datum
Avoid aligning on features of small area

5. SELECTION OF THE WORKPIECE ORIENTATION

Once the measurement and datum features have been determined the next step is to
decide on the orientation of the workpiece within the measurement volume of the CMM.
The major consideration is to ensure the accessibility of the surfaces and features that have
been selected for probing. As stated in section 3.2 the user should ideally seek a single set
up for the entire strategy.

When using a CMM the workpiece is not normally held on the datum features; it is these
features that require probing and they therefore need to be free of any obstruction.

The user should be aware of the functionality of the workpiece and in some cases it is
advisable to align the critical feature of length along one of the CMM's axes. Aligning the
artefact in this way will ensure that only one axis is used for measurement, therefore
eliminating the uncertainty due to the other axis.


17

6. SELECTION OF THE WORKPIECE HOLDING METHOD

6.1 MEASUREMENT FORCE

When choosing a holding method the user should be aware of the effect of the probing
force. CMMs operate at very low probing forces, usually in the region of 0.05 to 0.2 N.
However this is the force at the time of measurement. The CMM may subject the item to a
force of up to 3 N during the measurement process irrespective of the target measurement
force as many CMMs have an initial probing force far higher than the target measuring
force.

6.2 HEAVY ITEMS

If the part is sufficiently heavy then it may be possible for it to be located on the machine
table in a stable manner without the need for any holding devices. The user should note
that measurement in a free state, i.e. no use of securing medium of any kind, is now a
requirement under some geometric dimensioning and tolerancing standards (ASME
Y14.5M-1994). However be aware that heavy loads do distort the geometry of the
machine.

6.3 LIGHT ITEMS

For small, light parts plasticine, modelling clay or instrument wax can be used as a
holding material, although the user should ensure that after completion of the inspection
all traces of plasticine or clay are removed from the table and component. Magnetic or
vacuum chucks are alternative holding methods.

In some cases it will be necessary to mechanically clamp a workpiece to the CMM table or
jig. The user should be aware that clamping forces can distort the true shape of the part
and care should be taken not to over tighten the clamps. The use of cork sheet between the
workpiece and the clamps is advised. The Computer Numeric Control (CNC) program
should be written to avoid collision with the clamps and notes on the clamping locations
kept with the CNC program.

18

6.4 FIXTURES

Work- holding kits can be bought that enable the user to fabricate simplified, purpose-
built fixtures in which to locate workpieces.

For more complex set ups modular workpiece clamping systems, such as Allufix, should
be considered. Another example of good holding practice is to use purpose-made fixtures
for the part to provide ease of use for location; this workpiece holding technique has the
advantages of a short set up time and measurements that can be taken in one particular
part of the CMM table where the performance of the CMM is well known.

When using purpose-made fixtures the user should ensure that the fixture is clean, the
workpiece is located correctly, clamping forces do not distort the workpiece and that all
features to be measured are accessible.

6.5 ADHESIVES

Workpieces can also be secured directly to the table by the use of a glue gun containing an
appropriate glue stick for obvious reasons glues recommended for use on granite should
not be used on granite worktables. The component is secured by tacking the edge of the
component in direct contact with the table. The advantage of this method is that the
workpiece does not undergo distortion by clamping forces. The user should ensure that all
features to be measured are accessible; it must be borne in mind that it will not be possible
to gain access to the feature in direct contact with the table. When measurement is
completed the glue is removed by the application of a suitable solvent. The main
disadvantage of using this method is that the user will need to orientate the component by
visual means. The user must be aware that glue guns should only be used if:

(a) there is no significant change in the room temperature whilst the component is
glued to the table, or
(b) the expansion coefficient of the part being measured is close to that of the
worktable.

If the workpiece and granite temperatures increase, their lengths will increase at different
rates. Since the base of the part is glued to the table it cannot expand resulting in part
distortion. Clearly changes of temperature will lead to distortion of the part or the table if
the coefficient of thermal expansion of the part and table are significantly different.


19
Instrument wax is an alternative to glue and is heated and softened by hand and applied
to the edge of the workpiece in contact with the table in the same way as the glue is
applied. However since instrument wax can cause movement of the workpiece position
for about an hour after being applied, measurements should not be taken until this period
of time has eleapsed.

The use of restraint materials that are elastic by nature (such as silicon rubber) allow
expansion of the workpiece and are recommended.

6.6 GENERAL

In any clamping arrangement do not over-constrain. Ideally one would only clamp at one
point to minimise distortion, however, such clamping may allow the part to rotate during
measurement. Light clamping and the use of a material such as rubber or cork between the
workpiece and clamp will help to minimise distortion and also avoid damage to the
workpiece.

6.7 SUMMARY

Do not over constrain the workpiece
Do not clamp so as to distort or damage the workpiece
Be aware of thermal problems
Avoid the use of glue if possible

20

7. PROBE QUALIFICATION

7.1 BASIC QUALIFICATION

Probe qualification (calibration) is covered in a separate NPL good practice guide on
probing (No. 43). However the following is a brief description of some key points.

Probe qualification should be undertaken for every combination of stylus and probe body
that is likely to be used during the course of the inspection of a particular component. The
stylus tip should be qualified against a reference sphere (Figure 13) or other recognised
calibration artefact. It is most important to keep the stylus tip and qualification artefact
clean at all times. The user should adopt a task-related measuring strategy i.e. the
dimensions of the stylus tip should be qualified under the same conditions as those under
which the workpiece will be measured. Ideally the direction and speed of approach
should be the same for both qualification artefact and workpiece.

It is usual to use a certified reference sphere for qualification purposes. During
qualification the probe stylus is driven perpendicularly onto the surface of the sphere in
several planes. By qualifying the probe in suitable directions the effect of pre-travel
variation of a touch trigger probe over 360 can be significantly reduced.

Careful qualification of the stylus will improve the accuracy of some measurements, such
as features of size. By increasing the number of points used in the stylus tip calibration, the
resulting effective stylus tip diameter will generally be more accurate.

It is recommended that the smaller the diameter of stylus used the smaller the reference
sphere should be e.g. for qualification of a 1 mm diameter tip use an 8 mm diameter
qualification sphere. The user should also ensure that if more than one reference sphere is
available the value that is entered into the software definitely relates to the sphere in use.
A common error is the use of the wrong value for the reference sphere; if this happens
then the probing result output will be incorrect.

The user should maintain a written record of the historical data of probe calibration and
ensure that during probing of the reference sphere comparison is made with the data
being produced and that in the historical record. The question should be asked:

Does this probe normally output this value? If not why has the value changed.


21
If the values obtained from the probe qualification have altered significantly then the user
should take action to investigate the reason. The stylus and reference sphere should be
examined for cleanliness, and the user should ensure that the value entered into the
software corresponds to the reference sphere currently being used and that the probing
strategy is the same as that used to establish the historical data.

Figure 13 The probe configuration on a datum sphere for stylus tip calibration (courtesy F M M Chan et
al)

It is possible to use gauge blocks and ring gauges in order to qualify the probe stylus.
However the sphere is preferable since all probing directions are taken into account.

22

7.2 ADDITIONAL CHECKS

Gauge blocks can be used to determine the probe hysteresis of touch trigger probes - an
error that is a consequence of the direction of the previous trigger and following the
reseating of the probe. For analogue probes the test can detect errors in probe
qualification. The test is performed by probing the inner and outer faces of a gauge block
stack (Figure 14). The difference between the two readings should be zero; however any
difference in the actual results obtained will indicate the degree of uncertainty associated
with the particular probing system in use.

Figure 14 Probe contacts on gauge blocks and ring gauge (courtesy F M M Chan et al)
The gauge block check above will also reveal any errors in determining the radius of the
particular stylus in use. For example if the size of the middle gauge block is 10.000 0 mm
and the measured dimensions are 9.998 0 mm (external) and 10.002 0 mm (internal) this
suggests that the probe radius is incorrect by 0.000 5 mm.

If ME is the measured dimension externally, MI is the measured dimension internally, G is
the calibrated size of the gauge block, r is the actual probe tip radius and rq is the qualified
probe tip radius then

q E
r r G M 2 2 + =
q I
r r G M 2 2 + =
and
q I E
r r M M 4 4 =
The difference between the two measurements is four times the error in the probe radius.


23
The probing sphere radius will always make contact perpendicular to the surface and the
stylus should always approach the surface along this line. During measurement the probe
should be moved in the same direction as the measurements made during the qualification
stage. For example if a gauge block is mounted on the machine for measurement so that
the gauging faces are 45 to the X-axis then make sure that the qualification of the probe
includes probings in the same direction as will be used for probings on the gauge block
surface.

7.3 STYLUS CHOICE

The stylus tip diameter should be as large as possible; with larger balls surface finish has
less effect on the measurement being taken and more flexibility can be gained due to the
ball/stem clearance. The largest tip diameter that can be used is determined by the
smallest diameter hole to be measured (should the workpiece contain holes). Each stylus
has an Effective Working Length (EWL). This length is the penetration achieved by the tip
before the stylus shaft (stem) fouls on the component (Figure 15). Usually the larger the
stylus tip the greater the EWL.

Figure 15 (Effective Working Length source Renishaw Technical Specification Manual)

Stylus choice is covered in more detail in the CMM Probing Good Practice Guide No. 43.

24
7.4 COMPARISON (SUBSTITUTION METHOD)

The measurement of a traceable reference artefact will permit the errors associated with a
specific measurement task to be estimated. This allows the errors associated with the
parameters of a measured workpiece to be estimated directly. Because of the complex
error structure of the CMM, a calibration of this kind is only valid for measuring
workpieces that are nominally identical to the reference artefact used, measured in the
same location and using the same measurement strategy. Provided reference artefacts
(calibrated working standards) are available that are nominally identical to the workpiece
this type of strategy can achieve high accuracy with relatively little effort. This type of
strategy is referred to as the Comparator Principle and is relatively straightforward. The
reference artefact is measured several times by the CMM (it is good practice to make small
changes in the location of the object and the measurement strategy to establish the
sensitivity of the measurements to these effects). Once calibrated in this manner the
workpiece can be measured with a high degree of confidence. EAL document G17
Coordinate Measuring Machine Calibration covers this method in section 5.1 and Appendix
D. International standard ISO 15530-3:2000 also covers the use of a calibrated workpiece.
In particular section 7.4 covers the substitution method.

7.5 SUMMARY

Qualify the probe regularly
Perform additional checks against known artefacts
Pay attention to stylus choice
Consider the use of the substitution method

25

8. DEFINITION OF THE PROBING STRATEGY

8.1 INTRODUCTION

A machined workpiece is often a combination of standard geometric features such as
planes, circles, straight lines, cylinders, cones and spheres. The geometry of a feature can
be determined by the fitting of a substitute element to a number of measured points.

For measurement purposes there is a mathematically defined minimum number of contact
points (Figure 16) that must be used when fitting an element to a geometric feature. For
example two points define a line and three points define a circle. However using three
points for a circle will give no information on form. For practical purposes it is usual
during CMM operation to make more than this minimum number of points so that any
geometric error in the surface can be determined (Table 1). It is not necessary for the
points to be equally spaced over the surface but uniform coverage should be aimed for. In
fact it is recommended that equally spaced points are not used. Extreme distributions are
however not recommended.

The set of measured values made on the workpiece form the data input to the CMM
software in which calculations are carried out to determine position, size and departure
from nominal form. To obtain reliable results the data gathered should be representative
of the geometric feature being probed. Too few data points or data points that are
inappropriately distributed may provide an unreliable reference from which to determine
the result.

It must be borne in mind by the user that the greater the number of appropriately
distributed measured points the more reliable the assessment is likely to be. However the
larger the number of contact points used the longer the time needed to measure a feature.
It is necessary for the user to exercise economic judgement between the accuracy required
of the measurement process and the speed at which feedback of the results are required.


26

The following table is an extract from BS7172: 1989 and recommends the following
number of points per feature:
Table 1 number of contact points required
Number of contact points required
Geometric
Feature
Mathematical
Minimum

Recommended
Straight line 2 5
Plane 3 9 (Approximately three lines of
three)
Circle 3 7 (To detect up to six lobes)
Sphere 4 9 (Approximately three circles of
three in three parallel planes)
Cone

6

12 (Circles in four parallel planes
for information on straightness)
15 (Five points on each circle for
information on roundness)
Ellipse 4 12
Cylinder 5 12 (Circles in four parallel planes
for information on straightness)
15 (Five points on each circle for
information on roundness)
Cube 6 18 At least three per face

Figure 16 Defining common geometric features (courtesy Mitutoyo)

27

This guide cannot give complete advice on the measurement strategy to use because the
points chosen should take into account the nature of the machining process and the
intended function of the workpiece The user of the CMM may know the likely departures
from the nominal of a particular workpiece after a particular machining process. An
example of this is the type of finish imparted to a component being machined on a lathe
when using a round nosed turning tool with a heavy traverse. It may be that the finish on
the component could possess waviness that can affect the readings taken on the diameter.
The measurement procedure should take account of that knowledge. Section 8.2.2 covers
this idea in more detail.

The distribution of probing points should normally aim for a uniform coverage of the
feature being measured; this will ensure that the input data to the software provides a
genuine representation of the geometric feature. However the user should ensure that the
distribution should not be so regular that it has the potential to follow systematic or
periodic deformations that are a result of the machining process.

As an example of such deformation, if a nominal circle has three equally spaced lobes, a
distribution of six points equally spaced around the circle may fail completely to detect the
lobing effect. Thus, a certain amount of randomness in the distribution of points is
generally desirable. (Some software will automatically generate probing points for the
user in a CMM probing however these points are often generated to have a regular
distribution.)

During measurement the user should constantly question the results obtained and always
maintain an awareness of what is normal during a particular measurement strategy. If
something looks wrong then action should be taken and the suspect result investigated for
the possible introduction of errors during any step of the strategy. It is recommended that
records be kept of the output of all measurement programmes used and reference made to
them when performing repeat measuring strategies.

8.2 DISTRIBUTION OF POINTS

There are two basic approaches to choosing the number and distribution of points:
1. Apply ad hoc rules to generate probing points for the various fundamental
geometric features.
2. Measure and analyse in detail an actual component representative of the
components of interest in order to estimate the form error distribution over the part

28
and hence establish a probing strategy for subsequent use that provides an
acceptable balance between economy of measurement and accuracy of the result.
Approach 1 forms the basis of BS7172. It should be emphasised however that the rules
described in BS7172 are totally arbitrary and take no account of actual form error.

Approach 2 is scientifically based and is more suitable for parts coming off the production
line. A popular article describing this method appears in SSfM's Counting on IT, Issue 8
published by NPL.

For an isolated workpiece of which there is no a priori knowledge, Approach 1 can be
used, but there is no guarantee of the quality of the results. The detailed probing part of
Approach 2 would be relatively expensive, but reliable.

Section 8.2.1 covers the approach detailed in BS7172 whilst section 8.2.2 covers the
scientific approach.

8.2.1 The AD HOC approach

The approach detailed in BS7172 for various elements is summarised in the following
sections.

8.2.1.1 Lines

To achieve a nearly uniform distribution of points on a line segment, the line segment
should be divided into a number of sub-divisions of equal length and one point placed in
each subdivision

If the line is likely to suffer from a periodic distortion the chosen points should not
conform to a regular pattern. Therefore to ensure a non-uniform distribution a point
placed in a random position in each subdivision should be selected. To avoid the
possibility of choosing points too close together this process can be refined by dividing the
line into 3N-2 subdivisions of equal length and selecting a random point in each of the 1
st
,
4
th
, 7
th
,,(3N-2)
nd
sub intervals. Figure 17 shows the distribution of points on a line for
N = 5. One point is chosen at random in each subinterval marked with a asterisk.

* * * * *

Figure 17 A distribution of points on a line

29

8.2.1.2 Planes

To achieve a nearly uniform distribution of (N) points on a rectangular segment of a plane,
the rectangle should divided into a N
1
x N
2
sub rectangles where N
1
N
2
is approximately
equal to N. The sub-rectangles should be as near to a square as possible.

One point should then be placed in each rectangle. If the plane is suspected of suffering a
periodic distortion the chosen points should not conform to a regular pattern. A suitable
pattern can be achieved by selecting a random position for each individual point. If it is
more convenient to gather data on straight lines across the plane; then these should ideally
be irregularly spaced and the points on the lines selected according to the procedure
detailed above for line measurement. Figure 18 shows an example where N = 20: Choose
N1 = 4, N2 = 5 to give 4 x 5 sub-rectangles. Note: one point is chosen at random in every
sub-rectangle.

* *
* *
*
*
* *
* *
* *
*
* *
* *
* *

30
*
Figure 18 A distribution of points in a plane

If only a small number of points are to be measured the number of sub-rectangles should
be doubled and the points distributed in alternate sub-rectangles.

Figure 19 gives an example of this for N = 10: Choose N1 = 4, N2 = 5 to give 4 x 5 sub-
rectangles and chess board distribution of points. Note: one point is chosen at random in
every sub-rectangle marked with an asterisk.

* *

*

* *

*
*
*

*
*
Figure 19 A chess board distribution of points in a plane

8.2.1.3 Circles

To achieve a uniform distribution of N points the circle should be divided into N equal
arcs and one point placed on each arc. If the circle is likely to be lobed it is important not

31
to use a regular distribution. For example if it is known that there is a possibility of the
circle having q lobes, N should always be chosen so that N and q have no common factor.
(N should always be chosen to be greater than q.) If N were divisible by q then the
information gathered from the measurements would be severely limited.

As an example six points equally spaced on a 3 lobed circle may completely fail to detect
the lobing effect (see figure 13). A true circle fits the six points.
0.24
0.48
0.72
0.96
1.2
30
210
60
240
90
270
120
300
150
330
180 0

Figure 20 six uniformly spaced points (*) with complete failure to detect lobing

Seven points on such a circle will detect at least 79% of the amplitude of the lobing.

32
0.6
1.2
30
210
60
240
90
270
120
300
150
330
180 0

Figure 21 Seven uniformly spaced points (*) with at least 79% of the lobing detected

8.2.1.4 Spheres

The following strategy for determining the distribution of target points on a sphere
achieves a nearly uniform coverage with N points on the surface of the sector of the sphere
of radius r enclosed between two parallel planes that are distance h apart.

The procedure is:

First determine nc, an integer close to r Nh 2 / and then
np, an integer close to N/nc.

The standard then states that for each of the nc approximately uniformly spaced planes
parallel to and including the given planes, n
p
approximately uniformly spaced
measurements should be taken at (or near) the intersection of the plane and the sphere.

For a complete sphere, h = 2r, in which case nc is determined as an integer close to / N ,
with a single point at each pole.

For example:

Consider the case of a sphere where r = 100 mm, h = 150 mm and N = 30.

33

Using the formula r Nh 2 / determine nc as follows

nc = r Nh 2 /
nc = ) 100 2 /( 150 30
Therefore nc = 2.67 choose a value of nc of 3

np = N/nc,
np = 30/3
Therefore np = 10.00 choose a value of np of 10

Figure 22 illustrates the above example where r = 100 mm, h = 150 mm and N=30. Choose
nc=3 and np=10

Figure 22 A distribution of points on a sphere

8.2.1.5 Cylinders

To achieve a nearly uniform distribution of a number N of points on a cylinder of height h
and radius r is similar to producing target points on a rectangular plane segment as
detailed previously in the section on planes. The length of the plane segment will be h and
the breadth will be r 2 ; the distribution for such a plane may be used for the cylinder by
wrapping the plane around the cylinder.

A second method is to place the points on parallel circles, with the circles roughly
uniformly spaced.

nc, an integer close to r Nh 2 / should be determined along with np, an integer close to
N/nc.


34
nc is the approximate number of uniformly spaced planes approximately perpendicular to
the cylinder axis,

np is the approximate number of uniformly spaced measurements which should be taken
at the intersection of the plane and the cylinder.

For example:

Consider the case of a cylinder where r = 10 mm, h = 30 mm and N = 30

Using the formula r Nh 2 / to determine n
c
:
nc = r Nh 2 /
nc = ) 10 2 /( 30 30

Therefore nc = 3.78 choose a value of nc of 4

Using the formula N/nc to determine np:

np = N/ nc
np = 30/4
Therefore np = 7.5 choose a value of np of either 7 or 8 (see note below)

Figure 23 illustrates the above example

Figure 23 A distribution of points on a cylinder

35

Note: It is beneficial for the number of points to alternate between odd and even on the
circles, viz., in Figure 23 above seven points on the first circle, eight on the second,
seven on the third, etc. This choice will enable any lobing effect on the circular cross
section to be detected.

If the straightness of the cylinder is important then more circles should be used with fewer
points on each circle. If the circularity of the cross section is more important then more
points on each of a smaller number of circles should be used.

If the user is probing a helix then substitute circuit for circle in the above paragraphs.

Note: When measuring a hole whose axis is not square with the datum plane into which it
has been machined, the measurement strategy adopted by the user should be that
of treating the hole as a cylinder and not simply as a circle.

If all the touch points are taken in one plane, when the data are projected onto the XY
plane, the computed centre of the hole will not correspond to where the actual hole
intercepts the XY plane of the workpiece. The output data of the hole will indicate that it is
elliptical in nature.

To overcome the potential for a measurement error, the hole should be probed using the
method described above for a cylinder using planes parallel to the XY plane. Such probing
will enable the software to compute the position at which the axis of the cylinder intersects
the XY plane. The software will also compute the out-of-squareness of the hole in the XZ
and YZ planes.

36

Figure 24 Hole not square to datum plane XY ( D. R. Coleman & T. F. Waters)

37

8.2.1.6 Cones

A nearly uniform distribution of N points on the frustum of a cone can be produced in a
similar way as that on a cylinder with the target points placed on parallel circles. In the
case of a cone the number of points on the circles should be chosen to decrease towards
the vertex of the cone.

If the cone is of:
height h
side l
with radii r1 and r2 (r2 > r1 ) at its ends

The side length l is calculated using the formula l =
2
1 2
2
) ( r r h +

nc should be determined as an integer close to
) (
1 2
r r
lN
+

and s is determine as an integer close to l r r / ) ( 2
1 2

For each of the nc space planes that are approximately perpendicular to the cone axis, the
target points should be taken at the intersection of the plane and the cone. The number of
measurements on successive planes decreases by the integer s for a plane nearer the vertex
of the cone

Consider for example a cone where:

r1

= 10 mm,
r2

= 15 mm,
h = 20 mm,
N = 35:

Using the formula
2
1 2
2
) ( r r h l + = to determine the side l gives.

l =
2 2
) 10 15 ( 20 + ,
l = 20.6 mm,

Using the formula { } ) ( /
1 2
r r lN + to determine n
c
gives

38

n
c
= { } ) ( /
1 2
r r lN + ,
n
c
= { } ) 15 10 ( / 35 6 . 20 + ,
n
c
= 3.03 choose 3

Using the formula l r r / ) ( 2
1 2
to determine s

s = l r r / ) ( 2
1 2
,
s = 6 . 20 / ) 10 15 ( 2 ,

s = 1.53 choose 2

Figure 25 illustrates the above example which results in 10, 12 and 14 measurements (N =
36) on three circles.

Figure 25 A distribution of points on a cone

If the circularity of the cone is important, more points on the circles should be used.

If the straightness or angle of the cone is important, more circles should be used.

39

8.2.2 The scientific approach

A scientific method for choosing the distribution of points utilising prior knowledge of the
component form is described in the report Traceability of CMM Measurements: Influence of
the workpiece error on the measurement uncertainty (virtual workpiece) by Maurice Cox. Some
of the main points of this document are highlighted below. This section may be omitted at
first reading.

8.2.2.1 The problem

When a real feature on an industrial part is measured with a co-ordinate measuring
machine it is necessary to select a suitable probing strategy, i.e., the number and placement
of the measurement points. Once this decision has been made, the real feature is measured
according to this strategy. It is a difficult task to make this decision in an objective manner,
and particularly in a way that enables reliable estimates of the uncertainties of the
measurement results to be made. The decision is therefore often made in an arbitrary
manner. If a probing strategy (also known as a measurement strategy, measurement
procedure or sampling plan) is inadequate, an associated feature fitted to the measured
points can depart significantly from that which would be obtained from the use of a
superior probing strategy. The consequences of an inadequate probing strategy are that
the geometric parameters (diameter, angle, etc.) of the associated feature may differ
considerably from the correct values and therefore possess large uncertainties (which
cannot readily be quantified). Moreover, the estimate of the form error of the feature is
almost certain to be optimistically small. In many circumstances, it is expected that the
uncertainties due to the measurement strategy, as a consequence of form error, constitute
the major contribution to measurement uncertainty (Hocken, Raja and Babu, 1993;
Phillips, 1995; Weckenmann, Heinrichowski and Mordhurst, 1991). This point has also
been forcibly made at the International Standards Organisation Committee
TC/213/WG10, Coordinate Measuring Machines.

8.2.2.2 A solution

It is fundamental that in the absence of any prior information about the real feature being
measured it is not possible to provide a probing strategy that will ensure that the
measurement results will have a specified target uncertainty (or even to quantify the
uncertainty of the measurement results). However, if a particular feature, a cylinder say, is
being measured routinely as part of industrial production, it will be appropriate to learn
about that feature by making initially (and periodically thereafter) detailed measurements

40
of the part. From such measurements it is possible to deduce a probing strategy that is
economical for the target uncertainty and the effect of using it.

In taking detailed measurements, it is not essential that a CMM is used for this purpose,
even though subsequent measurements will for production purposes be made with a
CMM. Indeed, there can be advantages in using a different, special-purpose instrument
such as a form or a roundness-measuring instrument. For instance, a roundness-
measuring instrument can provide very detailed information about the profile of a section
of a real feature that is nominally cylindrical. Measurements should, however, be taken
with an instrument that is traceable and that operates to sufficient accuracy.

It is necessary that a probing strategy account for the uncertainty that results from the fact
that, even for the same number of points and the same relative placement of those points,
the measurement results will in general be different. As an illustration, consider the
measurement and re-measurement of a cylinder, before and after repositioning (e.g., a
rotation about its axis). Different points on the cylinder are measured in each instance,
thus sampling different regions of the surface (which contains form deviation), and
consequently different measurement results will be obtained. Another such situation
arises when it is necessary to decide how close to an edge of a real feature measurement
should be taken.

A considerable difficulty is the reliable estimation of these uncertainties. In particular, an
estimate of the maximum form deviation (MFD) of the real feature obtained from the
residuals of an associated feature will almost invariably be an underestimate, perhaps
significantly so, of the actual value, and hence be biased. This underestimation is a
consequence of the fact that only a (usually small) finite set of points is sampled. The bias
is in the unsafe direction, it generally being better for tolerancing purposes to determine, if
possible, an overestimate of this important quantity. Statements can also be made about the
parameters of the associated feature (and also about quantities derived from associated
features, such as the distance between the centres of two associated spheres,
corresponding to real features such as balls on a ball plate). The estimates obtained for a
given set of measurements could be underestimates or overestimates of the values that
would have been obtained from a detailed probing strategy.

Further, akin to the substitution method of measurement (ISO/TS 15530-3), the
comparison of previous detailed measurement with current coarse measurements
provides a basis for correcting the measurement results and the MFD in a way that bias is
reduced. Moreover, it also permits the uncertainties of these corrected values to be
estimated. In particular, uncertainties associated with either uncorrected or corrected
results can be obtained. To illustrate the point, the correction to the parameters and the

41
MFD obtained using a particular probing strategy, e.g., a probing strategy utilising 10
uniformly-spaced measurements of a nominally circular profile, is such that the values so
obtained correspond to (estimates of) those which would be obtained from a probing
strategy containing a large (and for practical purposes what amounts to an infinite)
number of uniformly-spaced measurements.

Finally, the approach provides a mechanism for designing optimal probing strategies.
Thus, questions such as What is the smallest number of measurements needed for a
probing strategy utilising uniformly-spaced measurements such that the uncertainty in a
particular (corrected or uncorrected) parameter is no greater than a specified amount?
can be considered.

The following assumption is made (this is expected to be a reasonable one in a range of
circumstances). The production process is such that machine-tool wear results in time to
gradual changes to the physical dimensions of the parts. On the other hand the nature of
the form deviation (which is typically a consequence of aspects of manufacturing-machine
performance other than tool wear) is such that once characterised it can be expected to be
substantially valid on a longer time scale. One of the purposes of regular inspection is of
course to ensure that changes to physical dimensions of a part and their effects are reliably
identified, in order that necessary adjustments to the production process can be made.

The methodology and procedures developed at NPL address the above issues. They are
based on the concept of a virtual feature. A representative manufactured part is synthesised
in terms of a mathematical model from which appropriate probing strategies can be
deduced. In particular, the approach provides an estimation of the uncertainties
associated with any particular measurement strategy.

Guidelines for assessing measurement uncertainty due to workpiece form errors are
overdue; work at NPL has made a contribution to the development of such guidelines. It
is expected that they will assist in assessing the contribution to the overall uncertainty
budget for the measurement of a part that arises from this consideration. Many of the
other contributions, viz., the uncertainties arising from the random and systematic errors
of measurement, can be estimated through the concept of the Virtual CMM, which was the
concern of the parent project of which this work was a part.

8.2.2.3 Basic principles and limitations

It is important to appreciate that it is not possible to realise an optimal probing strategy for
a complete class of features such as cylinders using any approach. An optimal probing
strategy depends on a number of factors including

42

1. the shape or aspect ratio (e.g., ratio of cylinder length to diameter),
2. whether the real feature is full (e.g., a complete sphere) or partial (e.g., a spherical cap),
3. accessibility to all or part of the surface of the real feature,
4. the form deviations of the real feature.

An optimal probing strategy can only apply to a particular class of closely related real
features, e.g., as indicated, to a sequence of nominally identical parts coming off the
production line (or to a set of parts that for other reasons are expected to be nominally
identical).

Assume that it is possible to take sufficient measurements on an initial part (or parts), and
periodically thereafter, to enable the parts of which it is representative to be adequately
characterised in terms of its form-deviation function and, consequently, its maximum form
deviation. For example, although parts may normally be measured with a probing
strategy containing, say, ten points, much more detailed information would be obtained
by measuring every 1000th such part, say, at one hundred points.

This information would be used as the basis of obtaining the detailed data needed for the
operation of the any method utilizing prior knowledge of the component form. It would
also be valuable for quality-control purposes. It is to be noted that, in the above very
simple example, only about 1% more measurements than normal would on average be
required. (This figure would be different in different circumstances.) However, there
would be gains in terms of the quality of the quantities estimated and their uncertainties,
with desirable consequences regarding the setting and meeting of manufacturing
tolerances.

The approach implicitly assumes in broad terms the constancy or slow changing of the
production process. However, it can also be used to assist in the monitoring of this aspect
of production.

The report mentioned at the start of this section shows how to construct and use such a
virtual feature in order to synthesise a real feature. By this means candidate probing
strategies can be studied, corresponding measurement errors deduced (if required to
provide corrections to measurement results) and optimal probing strategies can be
designed.

8.3 PARTIAL FEATURES AND PARTIAL ARCS


43
In general a partial feature is one that constitutes a fraction of a complete feature. Either
an actual feature might be of this form, or, because of access difficulties, it might be
possible to probe only part of a complete feature. Examples are an arc of a circle, a patch
of a sphere and a frustum of a cone. For instance, a partial (circular) arc might represent a
corner radius. Such features can be more difficult to measure than full features.

Because of the incomplete nature of the surface, errors can occur when trying to predict
the centre and radius of the best-fit circle from the co-ordinate data. It is often better to fit a
circle equal to the specified radius and then look at the deviations of form from this circle.

This point can be demonstrated by measuring a circular item of say 25 mm radius by
contacting the surface at 20 points around the circumference. Note the centre co-ordinates
and radius. Now make a measurement but this time contact at 20 points in a 45 sector of
the ring. Depending on the form deviations in the surface the results could be quite
different.

Consider a partial (circular) arc, measured at ten points uniformly spaced over part of a
nominally circular feature, as in Figure 26.

Figure 26 Ten uniformly spaced measurements on a partial arc, showing the (nominal) circle of which it
is part

Suppose a least-squares circle is fitted to these points to obtain the radius and centre co-
ordinates of the circle of which the arc is part. These circle parameters will have
uncertainties associated with them as a consequence of the CMM measurement
uncertainties. These uncertainties can be considerably greater when determined from
such partial arc data as opposed to the use of measurements giving sensible coverage of a
complete circle.


44
A partial arc will subtend a certain angle at the centre of the circle. Suppose the length of
an arc is halved, thus halving the subtended angle. Ten uniformly spaced measurements
taken as before, but over this shorter arc. The resulting uncertainty of the computed
radius is increased by a factor of approximately four, with a comparable statement
concerning the uncertainty of the centre co-ordinates. This result applies for circular arcs
that subtend any angle up to approximately 80.

The significance of this result can be seen by applying it to an arc subtending, say, 80, and
then one subtending, say, 5. The uncertainty in the radius determined for the latter case
is greater than that for the former by a factor of over 250.

In such extreme circumstances, it must be questioned whether the above circle parameters
are appropriate for an actual measurement task. To determine the radius and centre of a
partial arc to within a small uncertainty might require an accuracy of co-ordinate
measurement that is not readily available. A design specification in terms of such
quantities can be regarded as unreasonable. A specification that required the form
deviation (departure from circularity) to meet a certain tolerance would be much more
reasonable. Determining whether the measurements of the arc indicate acceptance in this
sense is simpler and forms a better approach. Information is available [*] on appropriate
approaches.

[*] G. T. Anthony, Helen M. Anthony, M. G. Cox, and A. B. Forbes. The parametrization of
geometric form. Technical Report EUR 13517 EN, Commission of the European Communities,
Luxembourg, 1991.

8.4 SUMMARY

Take care in the selection of the number and position of probing points
Too few points results may be in error
Too many points the measurement may take too long and errors will result from
drift due to changing conditions
Be aware of the problems associated with partial features


45

9. PROBING CONSIDERATIONS IN THE MEASUREMENT
STRATEGY

The performance characteristics of any probing system are influenced by the measurement
strategy selected by the user. For example the machine dynamic errors depend on factors
such as probing direction, probe speed, approach distance and acceleration values. The
effect of various probing parameters on the accuracy of a CMM fitted with a touch trigger
probe was investigated and reported in Some performance characteristics of a multi axis touch
trigger probe by F M M Chan et al (1997). Further information on this investigation can be
found in the NPL CMM Probing Good Practice Guide, No. 43.
10. PROGRAMMING THE CMM

Programming a CMM involves the use of a joystick, to instruct the machine on the route to
take to manoeuvre the probe around the workpiece, to define the number and distribution
of target points on each surface being measured and to define the computations to be
applied to the resultant data. The importance of this procedure should not be
underestimated since a program that contains an error could result in hundreds of parts
being measured incorrectly.

Errors can be reduced by producing a User requirement and Functional specification
before any programming is undertaken.

10.1 USER REQUIREMENT

The purpose of the user requirement document is to ensure that the person requesting the
measurements and the CMM programmer know exactly what is required. The user
requirement may be a simple statement such as Measure all features on the widget as per
drawing number 33333 using the datum detailed on the drawing. If no drawing exists
then a more detailed description may be needed. It is important in all cases that the target
uncertainty is defined. Trying to achieve 0.001 mm uncertainty when 0.005 mm is all that
is required is pointless. A form for providing a user requirement is included at
Appendix B.

10.2 FUNCTIONAL SPECIFICATION


46
The point of the functional specification is to allow the CMM programmer to plan his
measurement strategy so that errors and problems can be spotted early on. A functional
specification also allows for the program philosophy to be checked by another operator
before any coding has been done. Appendix C contains a suggested Functional
Specification form.

10.3 TESTING

The CNC program should always be tested before it is run in earnest. Appendix D gives a
suggested checklist.

One of the items is check measurements against those made with a rule. This may sound
ridiculous but many errors in programming a CMM may produce results that are wrong
by millimetres, e.g., by an amount equal to the probe radius.

More thorough testing of programs can be made using more traditional techniques such as
using micrometers, callipers, height gauges, etc. to measure a component that is then
measured on the CMM.

10.4 GENERAL

Many software packages contain an automatic feature measurement function that allows
the CMM to automatically move to a defined feature and take measurement points on it
facilitating the creation of a Direct Computer Controlled (DCC) CMM program. In such a
system the user defines the geometric feature by keying in its nominal geometric
information or by indicating the feature on a CAD model.

To obtain the most efficient use of the DCC CMM, particularly in high volume production
situations, the CMM should run part inspections as much as possible. The user should
ensure that programs are written off-line using a CAD model. Ideally the program should
be capable of being downloaded from an offline workstation into the CMM and run with
little or no editing.

The reproducibility of measurements is improved by using a DCC program. The use of
loops to measure the artefact more than once should be used and mean values calculated.
Conditional loops that allow a feature to be remeasured should form deviations be larger
than a set value should also be used. The downside of using a CNC program is that

47
exactly the same points are contacted each time. Deviations in form that lie between points
will be missed.

The user requirement, functional specification and testing forms mention earlier are
purely an aid to planning a CMM program. What is really important however is that a
mechanism for recording the information contained in the forms is available within the
organisation.

Records should be kept of all CNC programs and backup copies made. A revision history
that records the version number of the CNC program, details of any modifications made
to it and who made the modifications should be kept. The version number should always
appear on the printout from the program.

The key reason for this record is that it provides evidence of traceability. It should be
possible from the printout of the measurement of any component to be able to determine
at the very least:

The version of the CNC/DCC program used to make the measurements
The operator
The programmer
The date and time

10.5 SUMMARY

Plan the CNC program
Check the CNC program
Test the CNC program
Verify by another means

48

11. CMM DATA ASSESSMENT SOFTWARE

The data points that are recorded during the probing operation may contain two types of
error i.e. form deviation and measurement errors. These two errors can lead to errors in
assessment by the software.

Form deviation occurs in the workpiece itself and arises from generally repeatable aspects
of the manufacturing-machine performance such as machine tool elasticity and workpiece
fixturing. Consider, for example, data obtained by measuring a number of points on a
nominal circle on a workpiece. From the data gathered the software assessment would
determine an associated feature in the form of a best fitting mathematical circle and
compare its parameters with the corresponding values from the drawing. However if the
circle were to be measured using a different set of points form deviations would almost
certainly result in a different best fitting curve or surface.

Measurement error is on the other hand due to the inability of the co-ordinate-measuring
machine to provide exact values of the co-ordinates of points on the workpiece.

In practice the measured data can have errors that are either predominantly form
deviation or predominantly measurement error or it may contain errors of comparable
size from both sources. The way in which the software assessment of the data is
undertaken should take this into account.

Mathematical routines based on the principle of the least squares (Gaussian) method
assume that deviations from the nominal have a normal distribution. The result obtained
is the average size of the feature. However if the feature has significant departures from
the assumed nominal form the average size will be a misleading result. The alternative
method of analysis is based on the enveloping sizes method (Tchebyshev method).

49

Figure 27 Average size (Gaussian) and enveloping sizes (Chebyshev)( D. R. Coleman & T. F. Waters)

For example, if the measurement strategy used provides for a measuring accuracy that is
significantly greater than the accuracy of the machining process used to manufacture the
workpiece, the measurements can be taken as accurate information about the form of the
component. In this, case software assessment methods should be based on the largest
inscribed, smallest circumscribed or minimum zone forms.

However, if the opposite applies, i.e., the machining process accuracy is considered to
have much greater accuracy than the measuring accuracy, much of the form data will be
obscured by the measurement error. In this case, least squares analysis of the data is
recommended.

Software routines based on the Gaussian method give fast and consistent results whilst
routines based on the Tchebyshev method require advanced mathematical algorithms and
take longer to complete the analysis. For situations between these extremes, BS 7172: 1989
recommends that a combination of the least squares and other approaches be used. The
standard does not advocate a single approach to the analysis of data but urges the use of
methods that are appropriate to the task in hand.

Any software that carries out geometric form assessment should, in addition to the
required numerical output, provide the user with information on the method of analysis
used and a graphical output where possible. The output of a graphical record can often
allow errors resulting say from dirt on the component surface to be picked up quickly.

50

12. SOFTWARE FUNCTIONALITY

The wide variety of software available makes it impossible for this guide to give advice on
how to use the packages supplied with each and every machine. This section therefore
gives a general description of what is available and the functionality the user can expect
find.

CMM software has been developed and refined so that in general computer programming
knowledge is seldom required to run even the most sophisticated programs. The software
is usually intuitive, user-friendly and generally has a graphical interface that greatly
simplifies its interaction with CAD routines. Facilities such as reverse engineering,
reporting and editing tasks, provides the user with a wide set of capabilities for many
demanding applications.

The software normally consists of off-the-shelf, menu-driven programs and the user is
provided with a comprehensive array of help screens. Software routines are mostly easy to
use and can be customised to fit individual applications without having to use a specific
programming language.

The high level of compatibility of the software makes it easy to link CMMs and CAD
systems to create programs directly from the CAD data. Information can be imported as
either a wire frame solid or surface representation of the workpiece to be measured. CMM
software can also include automatic feature measurement from CAD data, best-fit
alignment, reverse engineering and 2D best-fit and 3D surfacing.

The interface between user and machine will often provide point-and-click measurement,
an interactive graphic display, alignment routines, both on- and off-line programming
capabilities, probe position display, and easy-to-read graphical form analysis.

The CMMs overall measurement capabilities and ease of use depend almost entirely on
the inspection software provided with the machine. The user should be aware of the
functionality offered by the software and aim to use it to its full capacity.

In summary the following features should be available:

Non-conventional part alignment
CAD interface
Graphical feature analysis
Automatic feature measurement

51
Reverse engineering
Offline programming
Programming language
Unknown geometry scanning
Geometric tolerancing
Complex surface measurement
Integrated best fit
Integrated statistics

These features are briefly explained below.

12.1 NON-CONVENTIAL PART ALIGNMENT

Most CMMs provide the ability to align planes, lines and points. But applications that
have offset datum planes or datum target points on complex surfaces are more difficult to
set up. Non-conventional part alignment software provides the facility to set up such
complex measuring applications.

12.2 CAD INTERFACE

CMM software may have the facility to accommodate multiple CAD file types and may
also present CAD information in a clear and understandable manner on the screen. It
should allow the user to control the graphical representation (wire frame, surface and
solid models) and to group and hide features for visual clarity.

12.3 GRAPHICAL FEATURE ANALYSIS

This software feature enables the user to evaluate the workpieces entire dimensional
condition allowing comparison of the touch points to the known CAD definition in
memory. Best practice in this area, for instance the finite probetip radius problem, is
considered beyond the scope of this guide but may form a topic for future guides.

12.4 AUTOMATIC FEATURE MEASUREMENT

This software facility can provide programming for Direct Computer Controlled (DCC)
CMMs. The CMM automatically moves to a defined feature and takes measurement
points on it. The user defines the geometric feature by inputting the nominal information
or by indicating the feature on the CAD model. Without this facility the user would need

52
to manually program the probe motion and the measurement points by use of the joystick
only.

12.5 REVERSE ENGINEERING

Using an Initial Graphics Exchange Specification (IGES) the CMM software may be able to
export measured data of an unknown or modified workpiece into a file format that can be
read by a CAD system to enable a CAD definition of the workpiece to be created.

IGES is a set of protocols for the transfer and display of graphical information on remote
devices via a telephone or computer communications network. IGES does not define any
new graphical file formats, but instead uses existing formats (such as Computer Graphics
Metafile) to encapsulate graphical data.

12.6 OFFLINE PROGRAMMING

To maximise the throughput from a DCC CMM, particularly in high production
operations, the machine should be used as efficiently and effectively as possible. In order
to avoid downtime when developing a measurement program it is essential that off-line
programming be utilised.

Offline programming software allows the user to write CMM measurement programs,
using CAD data, without interfering with the current measuring activities of the CMM.
The facility to download the program from an off-line workstation into the CMM and run
with the minimum of editing is an essential feature.

12.7 PROGRAMMING LANGUAGE

The CMM industry has its own industry standard inspection software language called the
Dimensional Measuring Interface Standard (DMIS). The use of this standard software will
allow interchangeability of part programs between different manufacturers brands of
DMIS-software. Some companies also use their own proprietary languages that allow
powerful yet easy-to-read programs to be developed.

12.8 UNKNOWN GEOMETRY SCANNING

This aspect of the CMM software offers the ability to scan complex features for data
without knowledge of its nominal definition. Generally the user will indicate the start and

53
end points of the surface and specify the scanning density. The CMM will scan between
the points gathering knowledge of the shape of the feature of surface. Best practice in this
area is considered beyond the scope of this guide but may form a topic for future guides.

12.9 GEOMETRIC TOLERANCING

All fundamental types of geometric tolerancing will probably be included in the software,
including line and surface profile.

It will normally be necessary to consider the uncertainty of measurement when
ascertaining compliance (or non-compliance) with a drawing or specification. UKAS
document M3003, Uncertainty and confidence in measurement, gives guidance in its
Appendix J. In addition an ILAC document on the same subject, viz., ILAC-G8:1996
Guidelines on Assessment and Reporting of Compliance with Specification (based on
measurements and tests in a laboratory), is available on the EA web site.

12.10 COMPLEX SURFACE MEASURING

The complexity of steps required for the alignment procedures for aerofoils, turbine blades
and other similar complex shaped products provides the user with a difficult challenge.
CMM software can provide the means to collect data on these complex products using a
CAD surface model of the part and should also allow the user to create section lines on the
surface to aid measurement.

12.11 INTEGRATED BEST FIT

Best fitting mathematically rotates and translates the part to a position that determines
whether the features as measured satisfy the design tolerances. This allows line and
surface profiles to be checked without a datum reference. The user can specify any
combination of features for inclusion in the best fit, enabling the results to be referred to
any Cartesian or Polar axis system. The software will allow the user to evaluate the entire
manufacturing process. Best fitting generally comes as third party software on an off-line
system, however integration into the CMM software improves the machines capabilities.

12.12 INTEGRATED STATISTICS


54
The ability to integrate statistical calculations into the inspection software eliminates the
need to transfer data into third-party software and enables statistical information to be
included in the assessment information record


55

13. ASSESSMENT INFORMATION RECORD

The user should ensure that an accurate record of the assessment software output is
available. This record normally has two parts - essential and optional. However the user
should add to this information details of the measurement strategy, e.g., probe, stylus,
extensions used, calibration artefact used to qualify the stylus, number of points used and
a brief description of the set up and direction of approach of the probe (see also section
10).

BS 7172:1989 details the following:

ESSENTIAL INFORMATION (SHOULD ALWAYS BE PROVIDED)

o Nature of assessment.
o Identification of the workpiece (e.g. serial number, type.)
o Details of the assessment (parameters of reference, departure from form etc.)

OPTIONAL INFORMATION

o Job title.
o Place, date and time.
o Environmental details.
o User.
o Any other information considered relevant.

Keeping records of this nature will provide the user with historical data that can be used
to reproduce a particular measurement scenario but more importantly will also establish a
knowledge base containing CMM inspecting experience that can be used for the
development of future measurement strategies.


56

14. ALIGNMENT CRITERIA FOR STANDARD FEATURES

This section describes the alignment criteria for some common measurement tasks.

14.1 LENGTH

The establishment of a datum for a workpiece in the form of a cylindrical shaft with planar
ends nominally orthogonal to its axis, for which a measurement of length is required, can
be performed in a number of ways.

The datum system can be established by sampling a minimum of five points along and
around the surface of the shaft. Using the Cartesian co-ordinate system the axis of the
cylinder is at the intercept of the XY and XZ planes. A further additional point is required
to define the YZ plane. The origin (X = 0, Y = 0 and Z = 0) is then set as shown in Figure
28.

One touch point at each end of the shaft is sufficient to determine its length. The software
will project the touch points onto the XY plane and surface algorithms and software
calculations will determine the length distance. In the case of a slot cut in the shaft parallel
to the axis of the shaft, the procedure to measure the length of the slot is the same as that
for the length of the shaft.

Figure 28 Two probing points for measuring a length ( D. R. Coleman & T. F. Waters)


57
A second method is to define a cylindrical polar co-ordinate system; in this case the three-
axis system will consist of two linear axes and one rotational axis. The Z-axis becomes the
axis of rotation and is perpendicular to the second linear axis. Once defined the shaft can
be sampled at either end and the distance calculated.

Figure 29 Cylindrical polar coordinate system ( D. R. Coleman & T. F. Waters)

A third method is to define the datum system by taking a minimum of three touch points
on one end of the shaft in order to define the YZ plane. Two points are taken along the
surface of the shaft to define the XZ plane and a further point on the cylindrical surface of
the shaft to complete the datum system. Once defined the shaft can be sampled at either
end and the distance calculated.

14.2 DIAMETER AND POSITION OF A HOLE

In this case the workpiece is a simple prismatic part with a cylindrical hole at right angles
to the XY plane (Figure 30). The diameter of the hole and the position of the centre line of
the hole relative to the origin are to be determined.

To define a datum system for a rectangular workpiece that has its origin in one corner
requires six touch points. Three points are required to define a flat surface or plane (the XY
plane), a further two points are required to define a line which lies on the XZ plane and a
final point defines the YX plane.

The three planes are mutually perpendicular and the point of intersection of these three
planes is the origin of the system, viz., X = 0, Y = 0 and Z = 0.

58

Having established the datum the next part of the strategy is to probe the hole; the
determination of the number and position of touch points to use can be calculated from
the section Distribution of points (8.2) in the guide. In this example the user is not
concerned with the hole depth, so the object can be treated as two-dimensional.

In two-dimensional measurement, the co-ordinates of the touch points are projected onto
the XY plane. The point co-ordinates are then used by the software to calculate the three
unknowns (hole diameter and distance of the hole centre line in the X and Y plane from
the origin).

An important point is that the hole diameter is determined by adding the diameter of the
stylus tip; this is always the case when probing a hole, the diameter of the stylus is always
subtracted when measuring a shaft. The user, by direct input, indicates whether a hole or
circle is being measured. Some systems can distinguish between these features by sensing
whether the probe moves away from the centre line after probing, in the case of a shaft, or
towards the centre line after probing as in the case of a hole.

Figure 30 Three probing points for measuring hole diameter ( D. R. Coleman & T. F. Waters)

The user should be aware that the measurement strategy adopted in this case is two
dimensional only. It the hole was not perpendicular to the XY plane then the result would
be incorrect. To obtain confidence in the result it is necessary to establish whether the hole
is perpendicular to the XY plane. To do this the hole should be sampled in accordance
with section 8.2.1.5 on Determination of points cylinders.

59

14.3 DIAMETER AND POSITION OF A SHAFT

In this case the workpiece is a simple prismatic part with a cylindrical shaft at right angles
to the XY plane (Figure 31). The diameter of the shaft and the position of the centre line of
the shaft to the origin are to be determined.

As described in section 14.2 the definition of a datum system for a rectangular workpiece
that has its origin in one corner requires six touch points. Having established the datum
the next part of the strategy is to probe the shaft; the determination of the number and
position of touch points to use can be calculated from section 8.2.1.3 Distribution of
points-circles in the guide. In this example the user is not concerned with the height of
the shaft, so the object can be treated as a two-dimensional problem when choosing the
measurement strategy.

In two-dimensional measurement, the co-ordinates of the touch points are projected onto
the XY plane. The co-ordinates are then used by the software to calculate the three
unknowns (shaft diameter and distance of the shaft centre line in the X and Y plane from
the origin).

An important point is that the shaft diameter is determined by subtracting the diameter of
the stylus tip, this is always the case when probing a shaft. The user, by direct input,
indicates whether a hole or circle is being measured. Some systems can distinguish
between these features by sensing whether the probe moves away from the centre line
after probing, in the case of a shaft, or towards the centre line after probing as in the case
of a hole.

60

Figure 31 Three points for measuring shaft diameter ( D. R. Coleman & T. F. Waters)

The user should be aware that the measurement strategy adopted in this case is two
dimensional only. It the shaft was not perpendicular to the XY plane then the result would
be incorrect. To obtain confidence in the result it is necessary to establish whether the shaft
is perpendicular to the XY plane. To do this it must be sampled in accordance with section
8.2.1.5 Determination of points cylinders.

14.4 AN ANGLE BETWEEN TWO LINES

The workpiece is a simple prismatic part with one corner cut away at an angle to the side
faces.

As described in section 14.2 the definition of a datum system for a rectangular workpiece
that has its origin in one corner requires six touch points. Having established the datum
the next part of the strategy is to probe the angled faces. This probing is carried out by
taking touch points on each face in order to define two lines. The determination of the
number and position of touch points to use can be calculated from section 8.2.1.1
Distribution of points-lines in the guide.

The software will compute the angle and normally provides the user with two angles, one
being the complement of the other; a decision will have to be taken as to which answer is
the correct one based on knowledge of the software.

61

Figure 32 Four probing points for simple angle measurement ( D. R. Coleman & T. F. Waters)


62

15. ENVIRONMENTAL CONDITIONS

The main environmental conditions that affect the CMMs accuracy and durability are
temperature, vibration and dust. For a CMM to be successfully used in a production line it
must be capable of operating to the required degree of accuracy in the environmental
conditions in which it is situated,

15.1 TEMPERATURE

The most significant factor that affects accuracy is temperature variation. Temperature
variations cause dimensional changes in both the workpiece and the CMM itself. An ideal
CMM would be completely immune to changes in temperature; however, it is practically
impossible to construct such an instrument and even if it were possible the component to
be measured would still be affected by changes in temperature. The next best would be a
CMM which would react to temperature in the same way as the component being
measured. However, with the diversity of engineering materials now in use this is also
impracticable. CMM manufacturers therefore minimise the structural distortion of the
CMM caused by the difference in thermal expansion coefficients of each component, by
employing a symmetrical design wherever possible. One manufacturer now uses hollow
granite for the structural members; these have low weight while retaining granites
favourable thermal characteristics. In addition, covers and bellows are normally used to
enclose the internal structural and moving members of the CMM therefore minimising the
effect of ambient temperatures changes. The user should regularly check that these covers
are in good condition. Points to note are:

Thermal installation should be used to enclose the inspection area or in the case of
in-line CMMs an enclosure should be used.
The air temperature surrounding the CMM should be controlled by the installation
of a heat exchanger or an air conditioning unit.
Care should be taken to maintain the proper temperature within the surrounding
area; a difference in temperature between the CMM and the workpiece can result in
measurement errors.
If the probing is commenced as soon as the workpiece is placed on the CMM table,
temperature changes during measurement may produce significant errors. The
workpiece should therefore be allowed to remain at the inspection temperature
until the user is satisfied that equalisation of temperature has taken place.


63
Compensation for thermal expansion and thermal distortion errors can be made by a
combination of hardware and software solutions known as adaptive structural thermal
compensation technology. The system includes a number of sensors capable of reading the
temperature on the structure of the machine and software algorithms that extrapolate
expansion and distortion values from the data. The software is able to compensate for the
current thermal state of the machine so that the influence of temperature variations is
virtually cancelled over a wide range. The resulting CMM is insensitive to shopfloor
temperature variations, as well as spatial and temporal thermal gradients. However for
best results care must still be taken to ensure that workpiece temperature effects are taken
into account.

The aim when installing a CMM should always be to keep thermal gradients to a
minimum. The second aim should be to minimise changes of temperature with time. The
final aim should be to operate the CMM at a temperature close to 20.0 C. If the machine is
not in a controlled environment then

Avoid opening or closing external doors and windows during a measurement
Do not turn on heating during the measurement
Shield the machine from heat sources (e.g. other machines)
Shield the machine from direct sunlight

15.2 VIBRATION

Vibration is a factor that can significantly affect the accuracy achieved by CMMs.
Machines used for in-line measurement must have sufficient resistance to vibrations
caused by machine tools and other equipment within the vicinity. Major design
considerations include a high natural frequency of the CMM in order to minimise the
effect of low frequency vibrations, rigid structural members and the use of air bearings
that require no contact between the bearing and the bearing surface during use, resulting
in zero wear and no noise or vibration.

In cases where the vibration levels are too high the user can improve the situation by
installing an appropriate vibration isolation system tailored to his production
environment.

The starting point is the information provided by the CMM manufacturer on the
maximum levels of vibration the machine can withstand. The next step is to relate this
information to the worst- case vibration characteristics on the production floor at the site
where the CMM is to be installed. Vibration on most production floors is usually the result

64
of either steady state or induced vibration or shock. A detailed knowledge of the nature
and magnitude of the vibration levels at the CMM installation site is therefore essential if
the CMM is to perform to specification. A vibration site survey can identify the
amplitudes of these vibrations and the frequencies at which they occur. Comparison of
this data with the manufacturer's allowable floor vibration limits then enables the correct
selection of an isolation system to reduce the transmission of floor vibration to within the
CMMs operating range. It is important that the correct equipment and data analysis
techniques are used in the site survey since errors here could seriously undermine the final
choice of isolation system.

The most commonly used isolators for CMMs, include elastomeric (rubber) pads, coil
springs and pneumatic isolators, and the information provided by the isolator
manufacturer on their vibration reduction characteristics should be studied before
implementing them on a CMM. It should be noted that foundations or inertia masses
might also be required in conjunction with the isolation system in order to increase system
stability or to provide additional stiffness and support for the machine bases.

15.3 DUST

Airborne dust from the production site can accumulate on the CMMs mating surfaces,
causing a slow deterioration in accuracy or possibly a system malfunction. For dust
protection, machines normally have built-in dust wipers on all axes; dust covers and
bellows protect the sliding scales. Regular maintenance will ensure that ingress of dust
and consequent wear of mating surfaces will be kept to a minimum. Cleaning of the
workpiece, stylus and calibration spheres should always be made with lint-free materials.

An effective measure to prevent dust from entering the working environment of the CMM
is the use of a fan to maintain an air pressure that is slightly higher within the area or
enclosure. The principle of this is that air will flow from the high-pressure area to the low-
pressure area thereby preventing dust from entering the atmosphere surrounding the
machine. Consideration should also be given to installing air filtration in the CMM
enclosure, the use of air-locks and TacMats and the wearing of suitable overalls and over
shoes by the operator since much of the dust can be fibres shed from unsuitable clothing.

15.4 SUMMARY

Beware of the effect of temperature
Keep all items clean
Take precautions to minimise dust

65
Select a low vibration installation site and take appropriate vibration isolation
measures

16. SUMMARY

The contents of this guide can be summarised as follows

choose appropriate datum
pay attention to the probe qualification
choose an appropriate number and distribution of points
beware of alignment and projection problems
beware of temperature effects
keep everything clean
beware of the problem of partial arcs
have in place adequate control of CNC programs


66

17. GLOSSARY OF TERMS

Accuracy For touch trigger probes the accuracy is stated in terms of the
uncertainty of measurement arising from error sources such
as repeatability and pretravel variation.

Analogue probe A proportional probe in which the displacement of the stylus
is represented by a probe continuously variable output
voltage or current proportional to the displacement.

CMM A measuring system with the means to move a probing
system and capability to determine spatial coordinates on a
workpiece surface.

Contacting probing system A probing system which needs material contact with a
surface being measured in order to function.

Probe The device that generates the signal(s) during probing.

Probing system A system consisting of a probe and, where present, probe
extensions, probe changing system, stylus, stylus changing
system and stylus extensions.

Probe head A device fitted to the ram of the CMM that carries the probe
mounting connector. Probe heads may have fixed orientation
or may articulate to provide re-orientation of the probing
axis. Articulating heads may be manually operated or
motorised.

Proportional probe A displacement measuring probe that provides an output,
which may be analogue or digital, proportional to stylus
displacement over a defined operating range.

Probe qualification The establishment of the parameters of a probing system
necessary for subsequent measurements.


67
Ram The component of a CMM that carries a probing system.
Also called the quill.

Requalification Repetition of the qualification procedure that may be
necessary after changing or moving components in the
measurement path, or following a change to ambient
temperature

Scanning Probe A proportional probe that is passed over the surface of the
workpiece in a continuous movement, sending data to the
processor at a high rate.

Shank A plain or tapered shaft for mounting a probe or probe head
to the quill of the CMM.

Stylus A mechanical device consisting of a stylus tip and shaft.

Stylus tip The physical element that establishes the contact with the
workpiece.

Touch trigger probe A discrete point-taking type of contact probe.

68

18. HEALTH AND SAFETY

Some optical trigger probes employ class 2 laser radiation therefore the appropriate laser
safety precautions should always be observed. Operators should wear laboratory coats or
overalls for safety reasons and to prevent fibres shed from clothing from falling on items
being measured.

Machines under direct computer control may move without warning. The operator should
stand back from the machine during an automatic run.

All should be aware of any regulations relating to the use of cleaning solvents (e.g.
COSHH)

69

APPENDIX A LINKS TO OTHER USEFUL SOURCES OF
INFORMATION

A.1 NATIONAL AND INTERNATIONAL ORGANISATIONS

A.1.1 National Physical Laboratory

The National Physical Laboratory (NPL) is a world-leading centre in the development and
application of highly accurate measurement techniques. As the UK's national standards
laboratory, NPL underpins the national measurement system (NMS), ensuring consistency
and traceability of measurements throughout the UK. NPL offers a unique range of
measurement services, contract research, consultancy and training services. Other areas of
expertise include the design and characterisation of engineering materials, and
mathematical software, especially its application to measurement and instrumentation.

For more information on the wide range of metrology services, facilities and research
work carried out at NPL either visit the NPL web site at http://www.npl.co.uk

Or contact the NPL Helpline at
Tel: 020 8943 6880, Fax: 020 8943 6458, E-mail: enquiry@npl.co.uk

A.1.2 National Institute of Science and Technology (NIST)

NIST is the equivalent of NPL in the United States of America. The NIST web site at
http://www.nist.gov/ often contains relevant documents in pdf format.

A.1.3 EUROMET, A European Collaboration in Measurement Standards

EUROMET is a cooperative organisation between the national metrology institutes in the
EU including the European Commission, EFTA and EU Accession States. The
participating metrology institutes collaborate in EUROMET, with the objective of
promoting the co-ordination of metrological activities and services with the purpose of
achieving higher efficiency.

The main aims of EUROMET are:
to encourage cooperation in the development of national standards and measuring
methods
to optimise the use of resources and services

70
to improve measurement facilities and make them accessible to all members

For more information visit the EUROMET web site at:
http://www.euromet.ch

A.1.4 European Co-operation for Accreditation (EA)

Until recently, the branches of European national accreditation bodies have been handled
separately by EAC (European Accreditation of Certification) and EAL (European co-
operation for Accreditation of Laboratories) dealing with certification bodies and
laboratories respectively.

These organisations have joined to form European Co-operation for Accreditation (EA)
that now covers all European conformity assessment activities including:
testing and calibration
inspection
certification of management systems
certification of products
certification of personnel
Environmental verification under the European Eco-Management and Audit
Scheme (EMAS) regulation

The members of EA are the nationally recognised accreditation bodies of the member
countries or the candidate countries, of the European Union and EFTA. Associate member
ship is open to nationally recognised accreditation bodies in countries in the European
geographical area who can demonstrate that they operate an accreditation system
compatible with EN45003 or ISO/IEC Guide 58.

The web site http://www.european-accreditation.org/ contains some useful information
and documentation on measuring techniques.

71

A.2 CLUBS

A.2.1 Dimensional Metrology Awareness Club (DMAC)

DMAC is an industry-focused club that has been set up as part of the 1999-2002 NMS
Length Programme. The club is managed by NPL on behalf of the DTI and has the
following objectives:
to provide a focused platform for interaction across all the technical areas covered
by the Length Programme
to enable members to learn about the latest developments in their own as well as
related areas of dimensional metrology
to provide a mechanism to encourage and facilitate interaction and exchange of
ideas with other member organisations
to give members the opportunity to provide input to, and influence the
development of future NMS programmes.
For further information visit the DMAC web site at:
http://www.npl.co.uk/npl/clubs/dmac/

A.2.2 Software Support for Metrology Programme (SSfM)

SSfM is a metrology-focused programme comprising projects designed to meet the
requirements of NMS programmes. The overall aim of the SSfM Programme is to tackle a
wide range of generic issues, some of which are problems in metrology that require the
application of established software engineering practices, whilst others require advances
in mathematics, software engineering or theoretical physics. The programme, thus,
includes work in metrology, mathematics, software and theoretical physics, with strong
links between the various disciplines.

The SSfM Club is aimed at users and suppliers of metrology software, giving them a say in
the direction of the Programme. It is the focal point for the distribution of general
information arising from the Programme.

Further details can be found at http://www.npl.co.uk/ssfm/


72
A.3 NATIONAL AND INTERNATIONAL STANDARDS

A.3.1 British Standards Institution (BSI)

BSI started in 1901 as a committee of engineers determined to standardise the number and
type of steel sections in order to make British manufacturers more efficient and
competitive. The BSI Group is now the oldest and arguably the most prestigious national
standards body in the world and is among the worlds leading commodity and product
testing organisations.

Further information on BSI can be found at:

http://www.bsi-global.com/group.xhtml

A.3.2 International Organisation for Standardisation (ISO)

The International Organization for Standardisation (ISO) is a worldwide federation of
national standards bodies from some 140 countries.

The mission of ISO is to promote the development of standardization and related activities
in the world with a view to facilitating the international exchange of goods and services,
and to developing cooperation in the spheres of intellectual, scientific, technological and
economic activity.

ISO's work results in international agreements that are published as International
Standards.

Further information on ISO can be found at:

http://www.iso.ch

The following BS and ISO specifications are relevant to this guide.

ISO 10360-2:1994 Coordinate MetrologyPart 2: Performance assessment of coordinate
measuring machines

BS 7172:1989 British Standard Guide to Assessment of position, size and departure from nominal
form of geometric features.


73
ISO 1101:1983 Technical drawings -- Geometrical tolerancing -- Tolerancing of form, orientation,
location and run-out -- Generalities, definitions, symbols, indications on drawings

ISO 1101:1983/Ext 1:1983 Toleranced characteristics and symbols -- Examples of indication and
interpretation (available in English only)

ISO 5459:1981, Technical drawings-Geometric Tolerancing - Datums and datum systems for
geometrical tolerances

ISO/TS 15530-3. Geometrical product specifications (GPS) Coordinate Measuring Machines
(CMM): Techniques for evaluation of the uncertainty of measurement Part 3: Use of calibrated
workpieces ISO International Technical specification, International Standards Organisation.

Document ASME Standard Y14.5M-1994 Dimensioning and Tolerancing may also be of
interest to users of this guide.

A.4 TRACEABILITY

Traceability in measurement is the concept of establishing a valid calibration of a
measuring instrument or measurement standard, by a step-by-step comparison with better
standards up to an accepted or specified standard. In general, the concept of traceability
implies eventual reference to an appropriate national or international standard.

The National Physical Laboratory is the United Kingdom's national standards laboratory.
It operates at the heart of the National Measurement System (NMS) which is the
infrastructure designed to ensure accuracy and consistency in every physical
measurement made in the UK. Chains of traceability link UK companies measurements
directly to national standards held at NPL.

For the majority of industrial applications, companies can establish a link to national
measurement standards through the calibration and testing services offered by UKAS-
accredited laboratories, which are in turn traceable to NPL. However, for challenging or
novel measurements to the highest standards of accuracy, which are not catered for by
UKAS-accredited laboratories, NPL can often provide a traceable measurement solution
directly to industry.

The United Kingdom Accreditation Service (UKAS) is the sole national body recognised
by government for the accreditation of testing and calibration laboratories, certification
and inspection bodies. A not-for-profit company, limited by guarantee, UKAS operates

74
under a Memorandum of Understanding with the government through the Department of
Trade and Industry.

UKAS accreditation demonstrates the integrity and competence of organisations
providing calibration, testing, and inspection and certification services.

Further information on UKAS can be found at http://www.ukas.com

A.5 NATIONAL MEASUREMENT PARTNERSHIP

The National Measurement Partnership programme is an initiative of the Department of
Trade and Industry with the aim of developing the organisational infrastructure of the
UK's National Measurement System.

The objectives of this programme are to:
increase the uptake of accredited calibration in the UK
increase the qualified skill base in measurement
increase user access to measurement expertise
To achieve these objectives, the programme is:
establishing a network of accredited UKAS Calibration Laboratories
expanding the Competing Precisely initiative to provide increased measurement
advice
establishing a National Measurement Helpline
managing the roll-out of measurement training in the form of a National Vocational
Qualification
establishing a National Measurement Forum to steer and support the programme

The National Measurement Partnership has produced a number of products to help
promote good measurement practice, including case histories, good practice guides and
self-assessment checklists.

For further details see http://www.nmpuk.co.uk


75

A.6 TRAINING COURSES

The Mitutoyo Institute of Metrology offers qualifications and training in over thirty
metrology related subjects. Mitutoyo training programmes are vocation based and
are accredited with the Open College Network (http://www.nocn.ac.uk) for a
qualification in Dimensional Metrology. These credits in turn, contribute towards
the evidence route of the Technical Services Engineering NVQ recently accredited
by EMTA (Engineering and Maritime Training Authority). These courses are
recognised nationally and are available in various areas of metrology.

See the Mitutoyo training pages http://www.mitutoyo.co.uk for more information.

The National Measurement Partnership through a DTI funded initiative has
developed supporting technical literature and provides good measurement practice
through National Vocational Qualifications (NVQ) in Technical Services
(measurement related evidence routes) for those who have a technical support role
(operators or semi-skilled employers). The programme is most suitable for people
who work to clearly defined procedures and are responsible for identifying and
implementing decisions.

Development and promotion of the NVQ with measurement routes was
undertaken by a consortium led by Brunel University, Coventry Technical College
and Mitutoyo (UK) Ltd. The activity was part of the National Measurement
Partnership programme managed by the National Physical Laboratory and funded
by the DTI as a project to support the National Measurement System.

See http://www.brunel.ac.uk/research/bcmm/measnvq.htm for information and
access to training manuals and CD ROM material on the Internet

Or visit: http://www.nmptraining.co.uk.

A.7 AWARDING BODIES

A wide variety of engineering courses offered by educational institutions include some
aspects of metrology studies, however the prospective student should enquire as to the
amount and content of dimensional metrology that is included in specific courses.


76
The following awarding bodies offer qualifications that can include aspects of metrology.
Information can be obtained from the awarding body by telephone or e-mail at the
following:

Edexcel http://www.edexcel.org.uk
Customer information centre: 0870 2409000
enquires@edexcel.org.uk

City & Guilds
Customer services: 020 7294 2800
Publication sales: 020 7294 2787
enquiry@city-and-guilds.co.uk

Oxford Cambridge and RSA Examinations
Customer services: 02476 470033
cib@ocr.org.uk

A.8 FURTHER READING

Measurement of Artefacts using repositioning methods, NPL Report CLM2 M G Cox, N R Cross,
A B Forbes and G N Peggs

Traceability of CMM Measurements: Influence of the workpiece error on the measurement
uncertainty (virtual workpiece) by Maurice Cox

Meas. Sci. Technol. 8 (1997) 837-848 Some performance characteristics of a multi-axis touch
trigger probe F M M Chan, E J Davis, T G King and K J Stout.

Precision Engineering 19:85-97, 1996 Error compensation for CMM touch trigger probes
W. Tyler Estler, S. D. Phillips et al.

Fundamentals of Touch Trigger Probing by David Coleman and Fred Waters
ISBN 0 9512010 1 8 Touch Trigger Press 1997

EAL G17 Coordinate Measuring Machine Calibration.

Fundamentals of Dimensional Metrology, Third Edition Ted Busch, Roger Harlow, Richard L.
Thompson.


77
Co-ordinate metrology Technology and Application by Han Joachim Neuman (Verlagmoderne
Industrie) Translation of Koordinaten Messtechnik by Ursula Brock.

Other information is published regularly in Quality Today, Measurement Science and
Technology, Metrologia, Measurement and Precision Engineering.
APPENDIX B USER REQUIREMENT

78

USER REQUIREMENT

Date:

Customer's Name:

Workpiece Description:
Drawing No:
Order/Internal OrderNo:

Date:
Uncertainty Required:
(E.g. 0.1 m)

Description of User Requirement (if customer order has insufficient information),
otherwise attach copy of Purchase Order.

Relevant British/International Standards:
(E.g. ISO 10360-2, 1994, etc.)


79
APPENDIX C FUNCTIONAL SPECIFICATION

80

FUNCTIONAL SPECIFICATION

1. Stylus Configuration

Probe Number
Length of Stylus (mm)
Diameter of Ball (mm)

Probe Number
Length of Stylus (mm)
Diameter of Ball (mm)

Probing Force:
Probing Speed:

2. Probe Calibration:
Manual
AP 473 (Semi-automatic)
NPL Special

Sketch of Probe.

3. Co-ordinate System:


81
4. Description of Probing Strategy:
Feature No. of Points Geometric Element


82

5. Temperature Correction:

Specified Ambient Temperature (C)

Coefficient of Expansion

6.
No. of Measuring Runs
Description of Programme:


83
7. Data Analysis:

8. Presentation of Results: Description of results to be output for generation of calibration
certificate

9. Data to be sent to mainframe? YES / NO

10.
Name Signature Date
Operator
Reviewed By


84
APPENDIX D TESTING

85

TESTING (CNC PROGRAMME)
CNC Programme Name:

Version No:
Written By:

Date Written:
Checked By:

Date Checked:

Check List Comment
1. Is the code consistent with
(a) User Requirement?

(b) Functional
Specification?

YES/NO

YES/NO

2. Are there any obvious errors in the
programme?
YES/NO
3. Hand check calculations (where
possible). Any inconsistencies?
YES/NO/
NA

4. Perform measurements of critical
features by another method (if possible).
Any inconsistencies?
YES/NO/
NA

5. Check all results against
measurements made with a ruler. Any
errors?
YES/NO
6. Run programme on an artefact
of known size (if possible).
Any inconsistencies?
YES/NO/
NA

MGPG 41

Uploaded by

Copyright:

Available Formats

MGPG 41

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MGPG 41

Uploaded by

Copyright:

Available Formats

CMM Measurement

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.