Th e I S O

m e t r i c a l
Geo
Pro d u c t
i f i c a t i o ns
Spec k
H a n d b o o

Find your way in GPS

ISO in brief
ISO is the International Organization for Standardization. ISO has a membership of some 162* national
standards bodies from countries large and small, industrialized, developing and in transition, in all regions of
the world. ISO’s portfolio of some 19 000* standards provides business, government and society with practical
tools for all three dimensions of sustainable development : economic, environmental and social.
ISO standards make a positive contribution to the world we live in. They facilitate trade, spread knowledge,
disseminate innovative advances in technology, and share good management and conformity assessment
practices.
ISO standards provide solutions and achieve benefits for almost all sectors of activity, including agriculture,
construction, mechanical engineering, manufacturing, distribution, transport, medical devices, information and
communication technologies, the environment, energy, quality management, conformity assessment and services.
ISO only develops standards for which there is a clear market requirement. The work is carried out by experts
in the subject drawn directly from the industrial, technical and business sectors that have identified the need
for the standard, and which subsequently put the standard to use. These experts may be joined by others
with relevant knowledge, such as representatives of government agencies, testing laboratories, consumer
associations and academia, and by international governmental and nongovernmental organizations.
An ISO International Standard represents a global consensus on the state of the art in the subject of that
standard.
* In December 2011

Danish Standards in brief

Danish Standards is the Danish national standards organisation. We offer our members influence on the
development of national and international standards.
Our services boost growth in Danish trade and industry and enhances quality in Danish society.
Our primary fields of activity include :
• Standardisation
• Consultancy
• Sale of standards and handbooks
• Courses and conferences
• Ecolabelling.
Through our subsidiary DS Certificering A/S we offer certification.
Within our core activity standardisation, Danish Standards concludes a performance contract with the Ministry
for Economic and Business Affairs. The contract sets out the framework and objectives for our activities as
a national standardisation organisation. A performance contract with the Danish Environmental Protection
Agency concerning the ecolabels, the Flower and the Swan, is also concluded.
Danish Standards is a private independent organisation. Our profits are ploughed back into the development
of new activities for the benefit of society and the corporate sector.
 The ISO
Geometrical Product Specifications
Handbook
Find your way in GPS

By
Henrik S. Nielsen
The ISO Geometrical Product Specifications Handbook. Find your way in GPS

January 2012
© ISO/Danish Standards 2012

Author: Henrik S. Nielsen

Illustrations: Henrik S. Nielsen

ISBN: 978-87-7310-721-8 (printed version)

ISBN: 978-87-7310-722-5 (pdf)
ISBN: 978-87-7310-723-2 (e-pub)

1. edition

This is a POD-publication
Printed in Denmark
Foreword

Geometrical product specifications or GPS is the international symbol lan-

guage that is used to express tolerances in technical drawings.
GPS is defined in a number of standards published by ISO, the International
Organization for Standardisation. The GPS standards are developed by ISO’s
technical committee TC 213. I am the chairman of that committee.
GPS is the foundation for manufacturing and international trade. GPS makes
it possible for a drawing of a component that has been developed in one
country to be sent to another country on the other side of the world, where
the drawing can be understood and the component manufactured, without
the designer and supplier having any common language except GPS.
The purpose of this book is partly to function as a text book in technical
schools and universities. However, it can also be used for self-study and as a
post-study reference.
The aim of this book is to give the reader sufficient knowledge to on one
hand read and interpret GPS drawings and on the other hand have enough
“vocabulary” and knowledge of the grammar to express geometrical require-
ments for a component as correctly formulated GPS requirements.
This book is a re-telling of the standards and has as such no legal standing.
There is no guarantee that there are no mistakes in this book. Further, because
the underlying standards are continuously developed and updated, even what
is correct in this book on the publication date can become wrong over time
as the standards evolve. Additionally, there are details and rules that apply in
certain situations that are not included in the book, partly to make the pre-
sentation less complicated and partly to limit the size of the book.
For all these reasons it cannot be emphasized enough that you always have to
go to the underlying standards, if you want to make sure you have the correct
interpretation of a GPS specification.
I would like to thank all who have helped make this book a reality. That in-
cludes Lars Brogaard of Danish Standards who first introduced the idea of an
ISO GPS book, Mikkel Hvass and the publishing staff at Danish Standards,

Find your way in GPS 3

 as well as the publishing staff at ISO who made this book a reality. I would
like to offer a special thanks to Mark Malburg, Ph.D. and his company Digital
Metrology Solutions, Inc. for letting me use test data and roughness diagrams
from his OmniSurf software package in chapter 9. Finally, I would like to
thank Per Bennich, Ph.D., the first chairman of ISO TC 213, for introducing
me to the GPS world in the first place.
I hope you will enjoy the book,
Henrik S Nielsen, Ph.D.
December, 2011.

4 Find your way in GPS

Contents

Foreword..........................................................................................................3
1 The GPS matrix system............................................................................9
1.1 The GPS philosophy.............................................................................. 9
1.2 Features................................................................................................. 10
1.3 The three worlds................................................................................... 13
1.4 The duality principle............................................................................ 14
1.5 Operations............................................................................................ 15
1.6 Operators.............................................................................................. 16
1.7 The GPS matrix system....................................................................... 17
1.8 Summary............................................................................................... 32
2 Dimensional tolerances.........................................................................35
2.1 Drawing indications............................................................................ 35
2.2 Types of dimensions............................................................................ 36
2.3 Features of size..................................................................................... 36
2.4 Two-point size...................................................................................... 38
2.5 Global size............................................................................................. 38
2.6 Envelope requirement......................................................................... 42
2.7 Indication of size requirements.......................................................... 44
2.8 ISO 286.................................................................................................. 45
2.9 Former practice.................................................................................... 48
2.10 Dimensions other than size................................................................ 49
2.11 Summary............................................................................................... 52
3 Tolerance zones......................................................................................55
3.1 3D tolerance zones............................................................................... 58
3.2 Set of 2D tolerance zones.................................................................... 65
3.3 Individual 2D tolerance zones........................................................... 70
3.4 Summary............................................................................................... 74
4 Tolerance symbols..................................................................................77
4.1 Fixed tolerance zones.......................................................................... 82
4.2 Mobile tolerance zones......................................................................100
4.3 Unrelated tolerance zones.................................................................128
4.4 Run-out tolerances............................................................................152
4.5 Summary.............................................................................................162

Find your way in GPS 5

 5 Datums and datum systems.................................................................163
5.1 The coordinate system of the component ......................................165
5.2 Datums and datum features.............................................................166
5.3 Locking of degrees of freedom.........................................................170
5.4 Datum systems...................................................................................172
5.5 Datum indicators...............................................................................176
5.6 Datum targets.....................................................................................177
5.7 Common datum.................................................................................179
5.8 Datums in the tolerance indicator...................................................182
5.9 Theoretically exact dimensions........................................................184
5.10 Summary.............................................................................................189
6 Combination of size and geometrical tolerances...............................191
6.1 Maximum material requirement.....................................................192
6.2 Maximum material requirement on datums..................................200
6.3 Maximum material requirement for TED patterns.......................212
6.4 Least material requirement...............................................................215
6.5 Summary.............................................................................................221
7 Special drawing indications................................................................225
7.1 All around...........................................................................................225
7.2 Common zone....................................................................................227
7.3 Projected tolerance zone...................................................................228
7.4 Flexible components..........................................................................229
7.5 Summary.............................................................................................230
8 General tolerances...............................................................................233
8.1 General dimensional tolerances.......................................................234
8.2 General geometrical tolerances........................................................236
8.3 Summary.............................................................................................238
9 Surface texture tolerances....................................................................239
9.1 What is surface texture?....................................................................240
9.2 Filtering...............................................................................................243
9.3 Basic parameters................................................................................248
9.4 Acceptance rules................................................................................255
9.5 Drawing indications..........................................................................257
9.6 Advanced parameters........................................................................266
9.7 Surface imperfections........................................................................270
9.8 Summary.............................................................................................273

6 Find your way in GPS

10 Edges.....................................................................................................275
10.1 Tolerancing of edges of undefined shape
according to ISO 13715.....................................................................278
10.2 Tolerancing of radii according to ISO 129-1..................................284
10.3 Tolerancing of chamfers according to ISO 129-1..........................286
10.4 Tolerancing of edges with geometrical tolerances.........................287
10.5 Summary.............................................................................................289
Appendix A – A tolerancing example.........................................................291
A.1 Establishment of datum systems......................................................299
A.2 Tolerancing of features of size..........................................................323
A.3 Positioning of features with fixed tolerances..................................325
A.4 Mobile related tolerances..................................................................333
A.5 Unrelated form tolerances................................................................336
A.6 Combination of tolerances ..............................................................338
A.7 Surface texture tolerances ................................................................345
A.8 Edge tolerances .................................................................................347
A.9 The finished drawings ......................................................................350
A.10 Non-functional features ...................................................................353
A.11 The expensive way ............................................................................356
A.12 Summary ............................................................................................361
Appendix B – The degrees of freedom indicator.......................................364
Index.............................................................................................................371
Word list.......................................................................................................375

Find your way in GPS 7

8 Find your way in GPS
1 The GPS matrix system

In this chapter we will look at the GPS matrix system, its structure and the
rules behind it. This chapter is quite theoretical and abstract and it can be
difficult to understand without at least a superficial knowledge of the GPS
system.
If you do not have this knowledge, you may consider skipping this chapter the
first time you read through the book and then return to it at the end. Hope-
fully the meaning of these rules will be clearer when you have read the rest of
the book.

1.1 The GPS philosophy

Every component has two types of characteristics: Material characteristics,
such as the strength and hardness of the material, and geometrical character-
istics. It is these two sets of characteristics together that control the functional
properties of the component.
GPS is an abbreviation of Geometrical Product Specifications. GPS controls
the geometrical characteristics of the component. The material characteris-
tics are controlled by other means. When we in the following talk about the
function of the component, what is meant is those aspects of the function that
depend on the component’s geometrical characteristics.
The basic ideas behind GPS are laid out in ISO/TS 17450-2 and are described
below.
The most basic idea in GPS is that the function of a component can be con-
trolled by one or more GPS specifications in the drawing. This means that the
designer can use the GPS language to control the function of the component
by expressing requirements for its geometry in the drawing.
All GPS requirements for GPS characteristics shall be expressed in the draw-
ing. There are no requirements that are implicit and which can be expected
to be fulfilled, if they are not expressed in the drawing. So it does not create
a perpendicularity requirement if two surfaces are drawn perpendicular to
each other, or a symmetry requirement if a set of holes are drawn symmetri-

Find your way in GPS 9

 cally around the centre line of the component, unless such a requirement is
explicitly written in the drawing.
The realisation of a GPS specification is independent of the specification it-
self. This means that the GPS specification defines requirements for the re-
sulting geometry, not for how the geometry is manufactured. So it is accept-
able (except in the few cases, where the manufacturing method is specified in
the surface texture specification), for a manufacturer to make the component,
so it fulfils the specification by milling and turning, where another manufac-
turer has to use grinding. This goes hand in hand with the idea that it is the
geometry of the component and not how it is made that controls the function
of the component.
The standard GPS rules and definitions for verification (measurement) are
formulated so they define theoretically perfect methods for the verification
of specifications. This does not mean that the verification has to follow these
rules and definitions to the letter. It means that what we call the measurand,
i.e. that which we are attempting to measure during the verification, is de-
fined in an unambiguous way. Consequently, there is one and only one true
value for each characteristic.
This does not mean that if the verification does not follow the rules it is wrong
and should be rejected. It means that there is measurement uncertainty as-
sociated with the verification and that this shall be taken into account when
proving whether a component fulfils a specification or not.
This further means that someone making a measurement has to make a tech-
nical and economical evaluation of how much measurement uncertainty is
acceptable compared to how expensive it is to make the measurement; how
expensive the component is; and what the consequences are of accepting a
component that does not fulfil the specification.

1.2 Features
Components consist of features in GPS. The subdivision of the component
into features takes place along the natural boundaries between the features,
i.e. usually along edges, see figure 1–1.

10 Find your way in GPS

 Figure 1–1:  The subdivision of a component into features

GPS is based upon this subdivision and GPS specifications express require-
ments for individual features, groups of features, or relations between fea-
tures.
There are situations where the boundary between two features does not create
an edge. Figure 1–2 shows a component that consists of three features, but has
no edges.

Figure 1–2:  A component without edges that consists of three

features: a cylinder, and two hemispheres, all of the same diameter

Find your way in GPS 11

 ISO 14660-1 defines a feature as a point, a line or a surface. As can be seen in
figure 1–3, a line can be a nominal (i.e. perfect) straight line (we often call a
straight line an axis) as the one shown in blue, a line with another nominal
shape (a curve) as the one shown in red, a closed line (e.g. a circle) as the one
shown in yellow or a real line (i.e. a line that has form error relative to its
nominal shape) as the one shown in green. Similarly, a surface can be a nomi-
nally plane surface (a plane) as the one shown in blue, a surface with another
nominal shape as the one shown in red, a closed surface (e.g. a cylinder) as
the one shown in yellow or a real surface (i.e. a surface that has form error
relative to its nominal shape) as the one shown in green.
Real lines and surfaces can have any nominal shape.

Figure 1–3:  A feature is a point, a line or a surface

12 Find your way in GPS

1.2.1 Feature types
ISO 14660-1 subdivides features into integral features and derived features.
An integral feature is a surface or a line in a surface on a component. You can
think of an integral feature as a feature you can touch.

Figure 1–4:  An integral feature is a surface or a line in a

surface

A derived feature is a centre point, a median line or a median surface, that is

derived from one or more integral features. A derived feature can be ideal, i.e.
without form error or non-ideal, i.e. with form error.

Figure 1–5:  A derived feature is a centre point,

a median line or a median surface

A feature of size is a special type of integral feature. It is a geometrical shape

defined by a linear size or an angular size. Features of size are either cylinders,
spheres, two parallel, opposite planes, cones or wedges. We will return to fea-
tures of size in chapter 2.

1.3 The three worlds

Components exist in three different worlds, or in three different models. The
first model is the nominal model, which exists in the CAD system or on the
drawing. The nominal model is a virtual (imaginary) model, where the com-
ponent has perfect geometry and no angular deviations.

Find your way in GPS 13

 The next model is the skin model, which is also a virtual model. However, the
component has geometrical and angular deviations in this model. The skin
model is a new model that was created as part of the GPS system. The speci-
fication is defined on the skin model. The advantage of defining the specifica-
tion on the skin model instead of the nominal model, as was the case before
GPS, is that the definitions for the specification take the geometrical and an-
gular deviations into account and remain unambiguous when the component
has such deviations.
The third model is the real model, where the manufactured physical compo-
nent exists. The verification of the manufactured component (of course) hap-
pens in the real world. The component is sampled during the verification, so
the real model consists of a set of points.
Figure 1–6 shows a component in the three worlds.

Figure 1–6:  A component shown as a nominal model (red), a skin

model (yellow) and a real model (blue), respectively

1.4 The duality principle

The duality principle is a concept that was introduced with GPS. The idea in
the duality principle is that a specification is expressed as a series of opera-
tions taking place in a given sequence, and that the verification mirrors these
operations. This mirroring is the duality in the duality principle.
This means that the specification is defined as a virtual measurement pro-
cedure. The advantage of this is that it makes it easier to determine if the
verification (the measurement) follows the specification and, where it differs,
this formulation makes it much easier to estimate the resulting measurement
uncertainty.

14 Find your way in GPS

1.5 Operations
When we say that the specification and the verification consist of operations,
we mean that they consist of a set of instructions, i.e. a kind of recipe for how
you get from the skin model or the real model to a measured value that can
be compared to the specification limits (the tolerance).
+, -, × and ÷ are examples of mathematical operations. GPS operations are
generally more complicated, but in principle they work the same way.
A mathematical recipe could be: Take a number, add 2, multiply the result
by 7, subtract 5, and divide the result by 3. This recipe has one result for each
possible starting number.
Similarly, a GPS specification recipe could be: Take a component, isolate a
certain feature of the component, extract a sufficient data set from the feature,
and filter this set such that all wavelengths in the surface that are longer than
0,8 mm are retained and all wavelengths in the surface that are shorter than
0,8 mm are excluded, derive a least-squares feature from the filtered feature,
find the largest positive deviation and the largest negative deviation between
the two features, and add these two values.
The GPS specification recipe in the example is much more complicated than
the mathematical recipe and it contains many terms and concepts that we
have not discussed yet, but the structure of the two recipes is the same: You
start with an input: A number for the mathematical recipe and a component
for the GPS specification recipe. After that you apply a number of operations
in a certain sequence until you arrive at a result that is a numeric value.
GPS has the following types of operations:
Partition separates the feature(s) involved in the specification.
Extraction defines a set of points that is the digital representation of a feature.
Filtration suppresses certain wavelengths in the surface.
Association defines an ideal feature (without form error) from a non-ideal
feature (with form error).
Collection considers a number of features as one entity.
Construction defines new ideal features from other features.
Evaluation defines a numeric value from one or more features. The evaluation
is always the last operation in a recipe.

Find your way in GPS 15

 1.6 Operators
What we have been calling a recipe is called an operator in both mathematics
and in GPS.
An operator is an ordered set of operations, which means that it describes a
number of operations that take place in a certain sequence.
A tolerance on a drawing defines what we call a specification operator and
one or two limit values. In other words, it defines a recipe for how you get to
a numeric value and gives either a limit that the value has to be smaller than
or a limit it has to be larger than, or both.
The specification operator applies to
the skin model and defines the “true”
measurement value, i.e. the value you Nominal model:
would get if you could make a perfect t mo-
measurement without measurement Theoretically perfec
it is
uncertainty. del of a component, as
CAD
represented e.g. in a
Any measurement we might make on system.
a real component has what we call a
verification operator built in. This
verification operator can to some
extent be described mathematically
when we use computer based mea-
suring equipment such as a coordinate measuring machine, but is often only
defined indirectly by the geometry and the working principle of the measure-
ment equipment. For example, if we measure the diameter of a cylinder with
a micrometer, we are measuring a local two-point diameter, which is defined
by the two highest points on the cylinder within the measuring anvils of the
micrometer. This definition is derived from the geometry of the micrometer
and its interaction with the geometry of the cylinder.
The verification operator works on the real component and defines the actual,
measured value. The actual, measured value has measurement uncertainty
from two different sources. The first is that the verification operator is not
identical to the specification operator. The second is that the verification op-
erator is not physically perfect. The first is called the method uncertainty; the
second is called the implementation uncertainty.
The method uncertainty is zero, if the verification operator in theory follows
the specification operator step by step. In this case the measurement uncer-

16 Find your way in GPS

tainty is made up entirely from implementation uncertainty, which occurs
due to the physical limitations in the measurement. If we look at the micro-
meter example above, the implementation uncertainty would include the flat-
ness and parallelism of the measuring anvils, inaccuracies in the measuring
spindle, the resolution of the micrometer, and temperature influences.
The difference between the verification operator and the specification opera-
tor leads to measurement uncertainty, if the former does not follow the latter.
If the specification for the cylinder in the example above specifies a require-
ment for the minimum circumscribed diameter (the diameter of the smallest
perfect cylindrical hole that the cylinder will fit into) and we measure the
cylinder with a micrometer, we will have method uncertainty because the
micrometer measures a two point diameter and not the minimum circum-
scribed diameter.
The magnitude of the difference
and thus the method uncertainty Skin model:
depends on the form error of the a com-
cylinder and cannot be predicted
Theoretical model of
form
without knowing it. But if we know ponent, which includes
e spe-
how the cylinder is made, we may and angular errors. Th
nent is
also know that it is sufficiently cification of a compo
odel.
straight, round and parallel for the defined on the skin m
difference to be so small that we
can live with it.
The idea in the duality principle is that instead of saying that there is one and
only one way we can measure a given specification and that all other measur-
ing methods are wrong, we can always make such an evaluation of whether
the verification operator is close enough to the specification operator and we
can use uncertainty to express this difference numerically.

1.7 The GPS matrix system

The GPS matrix system is defined in ISO/TR 14638. This system separates
the GPS standards into fundamental standards, global standards, general
standards and complementary standards. Table 1-1 shows the overall matrix
structure.

Find your way in GPS 17

 Global GPS standards
GPS standards or related standards that cover or influence several or
all GPS chains of standards
General GPS matrix (chains of standards)

1. Size 10. Location

2. Distance 11. Circular run-out
Fundamental GPS standards

3. Radius 12. Total run-out

4. Angle 13. Datums
5. Form of a line (unrelated) 14. Roughness profile
6. Form of a line (related) 15. Waviness profile
7. Form of a surface (unrelated) 16. Primary profile
8. Form of a surface (related) 17. Surface defects
9. Orientation 18. Edges

Complementary GPS matrix

Process specific tolerance standards
Machine element specific geometry standards

Table 1–1:  The GPS matrix system

18 Find your way in GPS

1.7.1 Fundamental GPS standards
The fundamental GPS standards are the most basic GPS standards. For the
time being only one fundamental GPS standard is published: ISO/TR 14638:
Geometrical product specification (GPS) – Masterplan.
Another fundamental GPS standard ISO 8015 (2nd edition) with the title
Geometrical Product Specifications (GPS) – Fundamentals – Concepts, prin-
ciples and rules has recently been published. Note that the 1st edition of ISO
8015 from 1985 is not considered a fundamental GPS standard.
ISO 8015 2nd edition describes what has until now been the unwritten rules
in GPS and geometrical tolerancing as it was known before GPS was intro-
duced.
We will cover these rules in the following.

1.7.1.1 The invocation principle

The invocation principle states that the entire ISO GPS system applies, if one
part of the ISO GPS system is used on a drawing. This means that all the
rules that are given in the following automatically apply without you having
to write it explicitly on the drawing. The purpose of the invocation principle
is to give customers and suppliers a rule to refer to, if the counterparty tries to
avoid following the GPS rules when a disagreement occurs regarding whether
a component is acceptable according to the specification.
The invocation principle means that you cannot disregard the rules that you
do not like if you use ISO symbols or ISO tolerance codes on a drawing, un-
less you explicitly write it either on the drawing or in the contract.

1.7.1.2 The principle of GPS standard hierarchy

The GPS standards are arranged in the following hierarchy:
1. Fundamental GPS standards
2. Global GPS standards
3. General GPS standards (the chains of standards)
4. Complementary GPS standards
The hierarchy principle states that rules in a standard at a higher level in the
hierarchy apply unless a standard at a lower level explicitly gives a different
rule.

Find your way in GPS 19

 For example, ISO 14253-1 gives global rules for proving conformance with a
specification, but the former edition of ISO 286-1 from 1988 in combination
with ISO/R 1938 from 1971 gave special rules when ISO tolerance codes for
size, such as H7 or b9, were used. This meant that the ISO 14253-1 rules did
not apply when ISO tolerance codes were used. Note that with the publication
of ISO 286-1:2010 and ISO 286-2:2010 this is no longer the case.
The ISO 14253-1 rules apply without you having to write it explicitly on the
drawing in situations where there are no special rules in a general standard or
a complementary standard.

1.7.1.3 The definitive drawing principle

The definitive drawing principle states that all requirements shall be indicated
on the drawing, in the documentation referenced on the drawing or in the
contract, and that it cannot be expected that requirements that are not indi-
cated will be fulfilled.
The typical situations where this is an issue are where surfaces are shown
perpendicular to each other on the drawing, or a number of holes are shown
symmetrically around the centre line of the component, without a tolerance
for the perpendicularity or the symmetry, respectively. In these cases it can-
not be expected that the perpendicularity or the symmetry meets a tolerance
that is not written on the drawing, just because it looks that way in the draw-
ing picture.
In this context it is useful to think of the drawing as consisting of three layers:
The first layer consists of the drawing medium, frames, title block, etc. This
layer does not create any requirements for the component and is not part of
the GPS system.
The second layer is the figures showing the drawing views that give the nomi-
nal geometry of the component. These views provide the skeleton that the
specification (the tolerances) is attached to. This layer does not create any
requirements for the component either and is not part of the GPS system.
The third layer is the tolerancing. It is only this layer that contains require-
ments for the component. It is this layer that the GPS system describes and
for which it creates rules.

20 Find your way in GPS

 Figure 1–7:  The drawing consists of three layers, the
drawing medium (bottom) with the frame and title
block, the views of the component (middle) and the
tolerancing (top)

There are a few situations that require a comment relative to this distinction.
There are cases where a reference is made to a given standard in or near the
title block. Such a reference is part of the tolerancing. Likewise, there can be
notes or references to e.g. company standards in the title block or elsewhere
on the drawing. These are also a part of the tolerancing and shall be regarded
as part of the third layer defining the requirements.

Find your way in GPS 21

 1.7.1.4 The feature principle
The feature principle states that the component consists of features with natu-
ral boundaries and that – unless otherwise specified – each GPS specification
applies to the entire indicated feature and only for this one feature.
A “natural boundary” is in most cases an edge, but there can be situations,
e.g. where a component consists of a cylinder with a hemisphere of the same
diameter at each end as in figure 1–2, where there are no real edges between
the three features. Another example is shown in figure 4–10 on page 91,
where a cam consists of two half cylinders and two plane surfaces and where
the cam profile therefore consists of four features.

1.7.1.5 The independency principle

The independency principle states that – unless otherwise specified – each GPS
specification for a feature or a relation between features shall be fulfilled inde-
pendently of other requirements in the specification.
There were some situations in ISO based tolerancing before the introduction
of GPS, where some types of tolerances were automatically linked together.
This is still the case in some national standards, e.g. the American ANSI/
ASME Y14.5.
The problem with this is that you as the designer unintentionally can change
the meaning of a specification (a tolerance) in one place on the drawing by
adding another specification somewhere else on the drawing.
It also forces the reader of the drawing to scrutinise the drawing in detail to
find any linked specifications there may exist in the drawing.
With the independency principle, the designer has to explicitly indicate if
two or more specifications shall be considered as one combined specification,
so the designer has full control over which specifications are combined and
which are not.
The independency principle also makes it possible to read and understand
the drawing one specification at a time, and that if two or more specifications
are combined, there will be symbols on the drawing to indicate this.
The independency principle makes it easier to construct the tolerancing in
the drawing and easier to understand it.

22 Find your way in GPS

Examples of characteristics that can be combined are size and form (see chap-
ter 2) and size and geometrical tolerances (see chapter 6).

1.7.1.6 The decimal principle

In GPS all numbers are considered exact. In other words, all non-indicated dec-
imals shall be considered as zeros and it makes no difference in the meaning
whether they are indicated or not. So ±0,2 is the same as ±0,200 000 000 000
and a deviation of 0,24 mm is not inside a ±0,2 m tolerance, even though you
would round down, if you only used one decimal.

1.7.1.7 The default principle

A default rule is a rule that applies when nothing else is specified. For ex-
ample, in some countries a speed limit of 50 km/h in urban areas is the default
rule. This applies everywhere in urban areas where a different speed limit is
not posted.
In the same way there are default rules in many areas of GPS that apply when
nothing else is specified. There are, for example, rules for the choice of cut-off
filter, tip radius, and sampling density that apply for surface roughness speci-
fications, when these details are not indicated explicitly.
A complete specification operator can be indicated with an ISO basic GPS
specification because these default rules exist.
A complete specification operator is a specification operator where all details
are defined. An ISO basic GPS specification is the shortest form for express-
ing a specification.
A basic GPS specification, which is defined in clause 3.5.4 of ISO/TS 17450-
2:2002 is the shortest form for expressing a specification, i.e. it is an indication
without modifiers that would change the default rules and the meaning of the
specification. An ISO basic GPS specification is a basic GPS specification that
is based on the rules in the ISO standards as opposed to e.g. the American
ASME standards or internal company standards.
The ISO basic GPS specification does not indicate that the details of the speci-
fication are undefined, but that the requirement is based on the default speci-
fication operator.
With the default principle a lack of modifiers does not mean that the choice
is left open during verification, but that the default defined in the GPS stan-
dards applies.

Find your way in GPS 23

 1.7.1.8 The reference condition principle
The reference condition principle defines the conditions under which the GPS
specifications apply for the component.
First and foremost all GPS specifications apply at 20° C according to ISO 1.
This is important because the vast majority of materials grow with increasing
temperature, and without a reference temperature size tolerances would be
meaningless.
Additionally, all requirements apply to clean components, therefore chips,
cutting fluids and other contaminants shall be removed before the compo-
nent can be verified correctly.

1.7.1.9 The rigid workpiece principle

All specifications apply to the component in the free state, i.e. without in-
fluence from external forces including the force of gravity. This means, for
example, that if the component is deformed by the measuring force, or by
gravity from its own mass, compensation shall be made for such deformation.
Tools that allow for tolerancing of non-rigid components and tolerancing
of components in a restrained condition are defined in ISO 10579:2010. We
cover this in chapter 7.

1.7.1.10 The functional control principle

As described above, GPS is based on the idea that the function of a compo-
nent only depends on the material properties and the geometrical properties
of the component.
Another way of formulating this idea is that any component function can be
expressed as what we call a functional operator. The functional operator de-
scribes the material properties and geometrical properties that are necessary
and sufficient to ensure that the component will function optimally.
The functional operator is a kind of virtual measurement that can predict
whether the component will function or not without fail. In reality it is, of
course, impossible to know the functional operator exactly.
The functional operator can in principle be formulated as a set of GPS speci-
fication operators for the geometric aspects of the functional requirements.
The task of the designer is to perform this translation of the functional opera-
tor to a necessary and sufficient set of specification operators.

24 Find your way in GPS

In this context “necessary” means that a specification operator has to be ful-
filled for the component to work satisfactorily. “Sufficient” means that all nec-
essary specification operators are present.
In other words, all necessary requirements shall be made to the component,
but no unnecessary requirements shall be made. If this is the case, all com-
ponents fulfilling all the specifications will function and all components not
fulfilling all the specifications will fail.
In reality, there are often either necessary requirements that are not expressed
in the drawing, or unnecessary or unnecessarily restrictive requirements in
the drawing.
A good designer is a designer who with as few unnecessary requirements as
possible can come as close as possible to describing all the necessary require-
ments for the component completely.

1.7.1.11 The general specification principle

We will cover general tolerances and how they are indicated on the drawing
in chapter 8. This is just a short description of the principle.
The general specification principle states that general tolerances only apply for
characteristics for which there is no individual specification. This means, for
example, that if general tolerances for dimensions are indicated on a drawing,
these general tolerances only apply to those dimensions where no individual
tolerance is indicated.
An individual specification can be more or less restrictive than the general
tolerance, so if there is a less restrictive tolerance indicated for an individual
dimension than the general tolerance for dimensions, it is only this less re-
strictive tolerance that applies to this dimension.

1.7.1.12 The responsibility principle

In GPS, specifications and verifications are not considered as either complete-
ly correct or completely wrong. Instead they are evaluated on their level of
uncertainty and/or ambiguity.
This uncertainty and ambiguity is then used to evaluate whether a specifica-
tion is good enough to describe a function and whether it is sufficiently un-
ambiguous to be used as the basis for the verification. See the chapters about
correlation ambiguity and specification ambiguity below.

Find your way in GPS 25

 Uncertainty and ambiguity is also used to evaluate whether a verification is
good enough to determine whether a component conforms to a specification
or not. See the chapters about measurement uncertainty and proving confor-
mance below.
The responsibility principle states that the designer is responsible for the uncer-
tainties and ambiguities associated with the specification, and that the party
proving conformance or non-conformance with a specification is responsible
for the uncertainties and ambiguities associated with the verification.

1.7.1.12.1 Correlation ambiguity

Correlation ambiguity is the ambiguity in the translation from function to
specification. This means that the correlation ambiguity quantifies how well
the designer succeeded in translating the functional requirements into the
specification.
Correlation ambiguity in the specification means that cases exist where a
component or a product fulfils the specification, but still does not work, or
that cases exist where a component or a product does not fulfil the specifica-
tion, but still works, or both.
It is very difficult, if not impossible, to quantify the correlation ambiguity
(which is why the term was changed from correlation uncertainty), but still it
is a useful concept that enables us to express the idea that there can be a dif-
ference between the functional requirements and the requirements expressed
in the specification.
The correlation ambiguity is the responsibility of the designer.

1.7.1.12.2 Specification ambiguity

The specification ambiguity expresses the ambiguity in the specification (!).
Specification ambiguity exists in cases where a specification can be interpret-
ed in several ways, either because the standards upon which the specification
is based are not unambiguous or because the designer has written the specifi-
cation in a way that allows for several reasonable interpretations.
In cases where the specification is ambiguous, it is up to the party that has
to show compliance with the specification to interpret the specification by
selecting among the reasonable interpretations. Figure 1–8 shows how this
ambiguity increases the tolerance interval for the specification. Figure 1–8
does not consider measurement uncertainty, which is discussed below.

26 Find your way in GPS

 Figure 1–8:  The specification ambiguity (Sa) increases the
tolerance interval

The designer is responsible for the specification ambiguity, just like the cor-
relation ambiguity.

1.7.1.12.3 Measurement uncertainty

The measurement uncertainty quantifies the differences between the specifica-
tion operator and the verification operator.
As described above, the measurement uncertainty can be split up into the
method uncertainty and the implementation uncertainty.
It is a general rule in GPS that the party proving conformance or non-confor-
mance with a specification is responsible for the measurement uncertainty.
The measurement uncertainty has to be subtracted from the tolerance inter-
val as shown in figure 1–9, if a supplier wants to show conformance with a
specification, i.e. prove that a component meets the specification and shall be
accepted. Figure 1–9 does not consider specification ambiguity.

Find your way in GPS 27

 Figure 1–9:  The measurement uncertainty shall be
subtracted from the tolerance when proving conformance
with a specification

The measurement uncertainty has to be added to the tolerance interval as

shown in figure 1–10, if a purchaser wants to show non-conformance with a
specification, i.e. prove that a component does not meet the specification and
shall be rejected. Figure 1–10 does not consider specification ambiguity.

Figure 1–10:  The measurement uncertainty shall be added to the

tolerance when proving non-conformance with a specification

28 Find your way in GPS

1.7.1.13 Proving conformance
If we look at the combination of the specification ambiguity, which is the
responsibility of the designer, and the measurement uncertainty, which is the
responsibility of the party proving conformance or non-conformance, we will
see two situations. One where the supplier wishes to prove that the compo-
nent meets the specification and shall be accepted, and one where the pur-
chaser wishes to prove that the component does not meet the specification
and shall be rejected.
The first situation is shown in figure 1–11. In this case the supplier can choose
any reasonable interpretation of the specification (if the specification is am-
biguous) and he only has to prove that the component fulfils one such inter-
pretation of the specification. The supplier then has to reduce the tolerance
interval by his measurement uncertainty to prove that the component meets
the specification.

Figure 1–11:  The specification ambiguity increases the tolerance

interval, and the measurement uncertainty reduces it, when the
supplier proves conformance with a specification

In reality, the net effect is that the tolerance interval has been expanded by the
specification ambiguity and reduced by the measurement uncertainty.
The other situation is shown in figure 1–12. In this case the purchaser has to
show that the component does not live up to even one reasonable interpreta-
tion of the specification (if the specification is ambiguous). The purchaser
then has to increase the tolerance interval by his measurement uncertainty to
prove that the component does not meet the specification.

Find your way in GPS 29

 Figure 1–12:  Both the specification ambiguity and the
measurement uncertainty increase the tolerance interval, when
the purchaser proves non-conformance with a specification

In reality, the net effect is that the tolerance interval has been expanded by
both the specification ambiguity and the measurement uncertainty.
The conclusion on these rules is that, as a purchaser, you shall always require
that the supplier provides measurement data to show that the purchased
components fulfil the specification. If you do that, you can use the functional
limits in your specification.
If you use incoming inspection instead, you have to reduce your tolerances in
your specification by the measurement uncertainty in your incoming inspec-
tion to ensure that you do not find yourself in a situation where you have to
accept non-functional components.

1.7.2 Global GPS standards

The next level in the hierarchy after the fundamental standards is the global
standards.
This level contains, among others:
ISO/IEC Guide 98-3: Uncertainty of measurement – Guide to the expression of
uncertainty in measurement (GUM)
ISO/IEC Guide 99: International vocabulary of metrology – Basic and general
concepts and associated terms (VIM)
ISO 1: Standard reference temperature for geometrical product specification
and verification
30 Find your way in GPS
ISO 10579: Dimensioning and tolerancing – Non-rigid parts
The ISO 14253 series: Inspection by measurement of workpieces and measuring
equipment
The ISO 14660 series: Geometrical features
The ISO 16610 series: Filtration
ISO 17450-1: Model for geometrical specification and verification
ISO 17450-2: Basic tenets, specifications, operators and uncertainties.
Except from ISO 10579 these standards do not contain any symbols that are
indicated directly on the drawing. They do contain global rules for how toler-
ances shall be interpreted as well as the entire concept universe that forms the
basis of the GPS system and GPS tolerancing.

1.7.3 General GPS standards

The general GPS standards make up the bulk of the GPS standards. These
standards contain the definitions of the characteristics that are toleranced for
the component’s features, as well as the symbols used in the tolerancing. It is
also these standards that contain the definitions for measurements, measure-
ment equipment and calibration of measurement equipment.
It is a fundamental idea in the GPS system that the general GPS standards are
organized in chains of standards.
A chain of standards consists of a number of related standards, that define
the tolerance symbol (chain link 1); the theoretical definition of the tolerance
(chain link 2); the definition of the tolerance on the real feature (chain link 3);
the measured value on the real feature (chain link 4); the metrological char-
acteristics for the measurement equipment (chain link 5); and the calibration
and verification of these metrological characteristics (chain link 6).
The underlying thought is that only when these six aspects of a tolerance
are unambiguously defined, it is possible to not only have an unambiguous
specification, but also a traceable verification for which you can estimate the
uncertainty and which therefore can form the basis for proving whether the
component meets the specification or not. Table 1-2 shows the general GPS
matrix. The GPS matrix system contains such a chain of standards for each
of the 18 characteristics listed in table 1-1. All the GPS standards contain an
annex that shows where the standard fits into the GPS matrix system.

Find your way in GPS 31

 Chain link
1 2 3 4 5 6

defined and measured

Drawing Definition of Definition of Measured Definition of Calibration

Comparison between
indication the tolerance the character- value of the metrological and verifica-
istic for the characteristic characteristics tion of metro-

characteristic
Codification Theoretical
definition of real feature for the real for measure- logical char-
the character- Specification feature ment equip- acteristics for
istic operator Verification ment measurement
operator equipment

Specification of GPS Verification of GPS

characteristics characteristics
Table 1–2:  The general GPS matrix

Practical use has shown that there is a break between chain links 3 and 4, such
that chain links 1-3 define an unambiguous specification and chain links 4-6
define the verification (the measured value). The break occurs because there
are some types of measurement equipment, for example coordinate measur-
ing machines, that can be used to verify many different geometrical char-
acteristics, so some chain link 4-6 standards cover more than one standard
chain and there can be several types of measurement equipment (and there-
fore several measurement equipment standards) that can measure the same
characteristic and cover the same standard chain.

1.7.4 Complementary GPS standards

We will not cover complementary GPS standards in this book.

1.8 Summary
In this chapter we have covered the GPS matrix system, its structure and the
rules behind it.
The functional properties of a component are controlled by the component’s
material characteristics and its geometrical characteristics. Geometrical
product specifications, GPS, can control the geometrical characteristics of the
component. The designer uses GPS requirements to control the function of
the component.
All GPS specifications shall be expressed in the drawing, and the realisation
of the GPS specifications are independent of the specifications.

32 Find your way in GPS

Components consist of features in GPS. The subdivision of the component
into features takes place along the natural boundaries between the features.
GPS specifications express requirements for individual features, groups of
features or for relations between features.
A feature is either a point, a line or a surface. There are both integral features
(real features that you can touch) and derived (virtual) features.
GPS is based on three different models: The nominal model, the skin model
and the real model. GPS specifications are based on the skin model and the
verification takes place on the real model.
The duality principle is a concept that was introduced with GPS. The idea in
the duality principle is that the specification is defined as a virtual measure-
ment and that the verification mirrors the specification.
GPS specifications consist of operators that are ordered sets of the following
operations:
"" Partition
"" Extraction
"" Filtration
"" Association
"" Collection
"" Construction
"" Evaluation.
The specification operator works on the skin model and defines the “true”
measurement value. The verification operator works on the real model.
Measurement uncertainty has two components: Method uncertainty that
characterises how well the verification operator follows the specification op-
erator, and implementation uncertainty that characterises the physical im-
perfections in the verification.
The GPS matrix system consists of:
""Fundamental GPS standards
""Global GPS standards
""General GPS standards (the GPS standard matrix)
""Complementary GPS standards.

Find your way in GPS 33

 The GPS system is based on the following rules and principles defined in ISO
8015, which is a fundamental GPS standard:
""The invocation principle
""The hierarchy principle
""The definitive drawing principle
""The feature principle
""The independency principle
""The decimal principle
""The default principle
""The reference condition principle
""The rigid workpiece principle
""The functional control principle
""The general specification principle
""The responsibility principle.
There are three uncertainties or ambiguities that characterise the differences
between the functional operator, the specification operator and the verifica-
tion operator.
The correlation ambiguity characterises the difference between the functional
operator and the specification operator. The specification ambiguity charac-
terises the ambiguity in the specification operator, and the measurement un-
certainty characterises the difference between the verification operator and
the specification operator.
When proving conformance with a specification, the specification ambiguity
shall be added to and the measurement uncertainty shall be subtracted from
the tolerance interval.
When proving non-conformance with a specification, both the specification
ambiguity and the measurement uncertainty shall be added to the tolerance
interval.
The global GPS standards contain the global rules for interpretation on toler-
ances and the entire concept universe that form the foundation of the GPS
system and GPS tolerancing.
The general GPS standards are arranged in 6 chain links in the general GPS
matrix. Chain links 1-3 define the specification operator and chain links 4-6
define the verification operator.

34 Find your way in GPS

2 Dimensional tolerances

Dimensional tolerancing is the classic way to tolerance drawings. The stan-

dards for dimensional tolerancing date back from the 1920’s and are in prin-
ciple unchanged since then.
In this chapter we will cover how dimensional tolerances are written on the
drawings, the possibilities they offer, and their limitations. We will discuss the
various types of dimensions you can tolerance, and when to use dimensional
vs. geometrical tolerances.

2.1 Drawing indications

The basic drawing indications for dimensional tolerances are defined in ISO
129-1.
+0,1 120,1
120 -0,2 120 +0,1/-0,2 119,8

Figure 2–1:  Indication of dimensional tolerances

Figure 2–1 shows the three ways to indicate that a dimension shall be be-
tween 119,8 and 120,1 mm. You can either write the nominal value followed
by the deviation limits with the upper deviation limit above the lower devia-
tion limit, or you can write the deviation limits on the same line as the nomi-
nal value with a slash between the values. The third possibility is to write the
limit values with the upper limit above the lower limit.

Figure 2–2:  Special dimensional tolerances

Figure 2–2 shows three special situations for dimensional tolerances. If one of
the deviation limits is 0, you have to write the zero. If the two limits are sym-
metrical around the nominal value, you can write ± followed by the allowed
deviation value, but you can also still use the indications in figure 2–1. Finally,
if a dimension is only a lower or upper limit, you can write the limit value fol-
lowed by min. for a lower limit or max. for an upper limit.

Find your way in GPS 35

 +0°0’30” 61°
60° -0°0’15” 60° ±0°0’15” 60° ±0,05° 59°

Figure 2–3:  Indication of angular tolerances

The same general rules apply to angular tolerances. Figure 2–3 shows a num-
ber of examples for angular tolerances. The only additional rule is that angles
and angular deviations can either be indicated as degrees, minutes (1/60 de-
gree) and seconds (1/60 minute or 1/3 600 degrees), with the symbols °, ’ and
” as shown in the first two examples, or with decimal degrees as shown in the
third example.

2.2 Types of dimensions

Dimensions can broadly be separated into those we call sizes and all the
others.
There are many rules and conventions that make size well defined. We will
cover those in this chapter.
There are very few rules for other types of dimensions. One of the reasons
for this is that it is very difficult to define a set of rules that are meaningful
for these dimensions in all contexts. This means that dimensions that are not
sizes are not well defined on the real component.
At the end of this chapter we will look at some of these problems and limita-
tions.

2.3 Features of size

Features of size are a special class of features that can be characterised by a
value, which is their size.

Figure 2–4:  Features of size with linear size

36 Find your way in GPS

Figure 2–4 shows the four types of features of size that have a linear size. This
group includes:
"" Cylindrical features, both outside cylinders (shafts) and inside cylin-
ders (holes), where the diameter is the size
"" Spheres, both outside and inside spheres, where the diameter is the
size
"" Two flat, parallel surfaces that face away from each other, where the
distance (the thickness, width or height) is the size
"" Two flat, parallel surfaces that face towards each other, where the dis-
tance (the gap) is the size

Figure 2–5:  Features of size with angular size

Figure 2–5 shows the two types of features of size that have an angular size.
This group includes wedges and cones, both of which can be either outside
or inside.
We make a distinction between local size, e.g. two-point size, which varies
from one location to another on the feature of size, if the feature is not geo-
metrically perfect, and global size, which is one value for the entire feature
of size. Minimum circumscribed diameter, i.e. the diameter of the smallest
perfect cylinder that fits outside the real cylinder, is an example of a global
size, but there are other global sizes. We will cover those later in this chapter.
Features of size are often part of a fit, for example when a shaft shall fit into a
hole. Many of the definitions and rules for size and size tolerances are focused
on ensuring that simple components can be assembled.
The vast majority of the features of size that are part of a fit are features of size
with linear size. As we are going to see, there are several possible rules and
definitions for these sizes, depending on how you indicate your tolerance.
Angular size on the other hand is not well defined at this time. There are spe-
cial standards for cones, but no general standards for angular size.

Find your way in GPS 37

 2.4 Two-point size
Two-point size is the simplest size definition. It is what we call a local size,
because it changes from one location to the next on the feature of size.
The two-point size, or two-point diameter, for a sphere or a cylinder is the
distance between two opposing points that are chosen such that the line be-
tween them goes through the centre. For a cylinder, the line also has to be
perpendicular to the axis.
The local two-point size for a feature of size consisting of two parallel, oppos-
ing planes is the distance between two points that are chosen such that the
line between them is perpendicular to the median plane.
You often specify requirements for the largest two-point size for holes and in-
side gaps, such as the fourth picture in figure 2–4, and the smallest two-point
size for shafts and blocks, such as the third picture in figure 2–4, for example
to ensure the strength of the component or the uniformity of fit, so it does not
become too loose, which makes the features move about and damage each
other.

2.5 Global size

A global size, as opposed to a local size, is a size that can be described by a
single value for the entire feature. The most important global sizes are maxi-
mum inscribed size, minimum circumscribed size and least-squares size.

2.5.1 Maximum inscribed size

Maximum inscribed size is normally used for internal features, i.e. cylindrical
holes and gaps, such as the fourth picture in figure 2–4.
For a hole, the maximum inscribed diameter is defined as the diameter of the
largest perfect cylinder that can fit into the hole in its entire length, see figure
2–6. The maximum inscribed diameter defines the largest shaft that can fit
into the hole and is therefore important for fits.

38 Find your way in GPS

 Figure 2–6:  Maximum inscribed diameter

For a gap, the maximum inscribed size is the distance between two perfect,
parallel planes that are as far apart as possible while still fitting in the gap, see
figure 2–7.

Figure 2–7:  Maximum inscribed size

2.5.2 Minimum circumscribed size

Minimum circumscribed size is normally used for external features, i.e. shafts,
spheres and blocks, such as the third picture in figure 2–4.
For a shaft, the minimum circumscribed diameter is defined as the diameter
of the smallest perfect hollow cylinder that fits outside the shaft in its en-
tire length, see figure 2–8. The minimum circumscribed diameter defines the
smallest perfect hole the shaft will fit into and is therefore important for fits.

Find your way in GPS 39

 Figure 2–8:  Minimum circumscribed diameter

For a sphere, the minimum circumscribed diameter is defined as the dia-

meter of the smallest perfect sphere that will fit around the real sphere.
For a block, the minimum circumscribed size is defined as the distance be-
tween two perfect, parallel planes that are as close to each other as possible,
while still fitting around the block, see figure 2–9.

Figure 2–9:  Minimum circumscribed size

2.5.3 Least-squares size

The least-squares size is an average size for the entire feature.
For a cylinder (a shaft or a hole), the least-squares diameter is the diameter
of a perfect cylinder that has a size such that the square of the outward and
inward deviations between the shaft or the hole and the least-squares cylinder
is as small as possible, see figure 2–10.

40 Find your way in GPS

 -

Figure 2–10:  Least-squares diameter

For a sphere, the least-squares diameter is the diameter of a perfect sphere

that has a size such that the square of the outward and inward deviations
between the real sphere and the least-squares sphere is as small as possible.
For a block or a gap, the least-squares size is equal to the distance between
two perfect, parallel planes that have a distance such that the square of the
outward and inward deviations between the two surfaces that define the
block or the gap and the two perfect, parallel planes is as small as possible,
see figure 2–11.

-
-

Figure 2–11:  Least-squares size

Find your way in GPS 41

 2.6 Envelope requirement
The envelope requirement simulates fit and is designed to ensure assembla-
bility of simple components such as shafts and holes.
The envelope requirement means that the feature is not allowed to violate a
virtual feature of perfect form at the maximum material limit. This means
that a shaft shall fit into a perfect, cylindrical hole with a diameter equal to
the upper tolerance limit. The shaft shall fit into the hole over its entire length.
In other words, the upper tolerance limit limits the minimum circumscribed
diameter of the shaft.
For a hole, the maximum material limit is the lower tolerance limit (there is
most material when the hole is the smallest). This means that a perfect, cy-
lindrical shaft with a diameter equal to the lower tolerance limit shall fit into
the hole. The shaft shall fit into the hole over its entire length. In other words,
the lower tolerance limit limits the maximum inscribed diameter of the hole.

Feature of size:
on,
ur e th at is ch ar ac terised by a dimensi
Feat e (OD
ic h is it s si ze . Fe atures of size includ
wh pla-
res, pairs of opposite
and ID) cylinders, sphe linders, spheres, and
ne s an d we dg es . Cy
nes, co ze
of op po si te pl anes have a linear size (a si
pairs angu-
m ill im et re s) . Co ne s and wedges have an
in
ees).
lar size (a size in degr

42 Find your way in GPS

d L d U

Figure 2–12:  Envelope requirement

As you can see, these two requirements together ensure that a shaft and a hole
will just fit together, if the upper tolerance limit for the shaft is equal to the
lower tolerance limit for the hole.

Find your way in GPS 43

 2.7 Indication of size requirements
The main standard for the definition of the meaning of size specifications in a
drawing is ISO 14405-1 that was published in 2010. According to ISO 14405-
1 the default meaning of size tolerances is the two-point size, i.e. the largest
two-point size has to be smaller than the upper tolerance limit, and the small-
est two-point size has to be larger than the lower tolerance limit. ISO 14405-1
defines a number of modifiers that can be indicated to change the definition
of the size requirement e.g. to an envelope requirement.
There are currently two ways to indicate size specifications in a drawing.
The first method is to indicate numeric values as shown in figure 2–1 and 2–2.
The other method is to use ISO 286-1 tolerance codes. There is no difference
in the meaning between the two. Both specifications create two-point size
requirements. We will discuss “ISO tolerance codes” as they are often called,
later in this chapter.
You modify the specification to indicate the envelope requirement by adding
a modifier after the tolerance as shown in figure 2–13. The indication is
the same regardless of whether it is an outside or inside size, but as we have
seen, the envelope requirement always applies to the maximum material lim-
it. In other words, the envelope requirement applies at the maximum material
tolerance limit as discussed in 2.6, and the two-point size still applies at the
minimum material limit, i.e. the smallest two-point diameter shall be larger
than the lower tolerance limit for a shaft, and the largest two-point diameter
shall be smaller than the upper tolerance limit for a hole.

Figure 2–13:  Indication of envelope requirement

The reason for these default definitions is that usually there are many size
tolerances in a drawing and often only a small fraction of these is part of a fit.
Envelope requirements are more expensive to manufacture and verify than
two-point requirements and are only necessary for fits.

44 Find your way in GPS

So you have the tools available to indicate which size tolerances have to work
as part of a fit and which ones do not. This makes the drawing more precise
and the manufacture of the component cheaper.

2.8 ISO 286

ISO 286-1 defines the ISO tolerance code system for holes and shafts. Each
code consists of one or two letters and a number. The letter (only one letter is
used in the vast majority of cases) indicates the fundamental deviation for the
tolerance, i.e. where the tolerance limit closest to the nominal value is located
relative to the nominal value.
The number indicates the
tolerance grade or the
width of the tolerance.
The number codes are ISO code letter case:
called IT grades. IT means
to be assembled”.
“International Tolerance” “Holes shall be large
rge) letters are
and is typically referred to Therefore capital (la
fundamental de-
as tolerance grade. used to indicate the
The values for the ba- viation.
all to be assemb-
sic deviations and the “Shafts shall be sm
case (small) let-
IT grades vary with the led”. Therefore lower
dicate the funda-
nominal diameter of the ters are used to in
feature. mental deviation.
The idea is on one hand
that a given tolerance
grade, e.g. IT7, can be
made with the same man-
ufacturing process for all
diameters and on the other hand that a fit consisting of e.g. a f6 shaft in a
H7 hole has the same character and feels equally tight or loose, regardless of
whether the diameter is 10 mm or 1 000 mm.
In reality it does not quite work this way, because the tolerance for a given tol-
erance grade does not change proportionally to the diameter. This means that
it is normally harder to achieve a certain tolerance grade for a large diameter
than it is for a small diameter, and a given fit combination will be tighter and
have less variation for a larger diameter.

Find your way in GPS 45

 There are two systems for tolerancing with ISO code:
If you use hole basis, you specify all holes using the same tolerance code, typi-
cally H7 or H6. You then adjust the fits by varying the shaft tolerances. The
advantage of using hole basis is that it limits the number of plug gages that are
necessary to be able to verify all the holes with envelope requirements.
If you use shaft basis, you specify all shafts with the same tolerance code,
typically h7 or h6. You then adjust the fits by varying the hole tolerances. The
advantage of using shaft basis is that you can use standard shafts and pins that
can be bought off the shelf.
The following applies for the letter codes for hole tolerances:
"" A through H tolerances are above the nominal size and an h shaft will
fit into the hole without interference
"" K through ZC tolerances are below the nominal size and an h shaft
will have to be pressed into the hole and will create an interference fit
"" H tolerances have the lower tolerance limit equal to the nominal size
"" JS tolerances are symmetrical around the nominal size
"" For A through H tolerances, the letter defines the lower limit of the
tolerance
"" For K through ZC tolerances, the letter defines the upper limit of the
tolerance.
The following applies for the letter codes for shaft tolerances:
"" a through h tolerances are below the nominal size and the shaft will
fit into an H hole without interference
"" k through zc tolerances are above the nominal size and the shaft will
have to be pressed into an H hole and will create an interference fit
"" h tolerances have the upper tolerance limit equal to the nominal size
"" js tolerances are symmetrical around the nominal size
"" For a through h tolerances, the letter defines the upper limit of the
tolerance
"" For k through zc tolerances, the letter defines the lower limit of the
tolerance.
The letter always defines the limit that is the closest to the nominal size for
both holes and shafts.

46 Find your way in GPS

Example: Shaft tolerance 50 mm c8
You find the shaft tolerance for a 50 mm c8 tolerance in ISO 286-1:2010
as follows:
First you find the fundamental deviation in table 4. For a 50 mm c toler-
ance the fundamental deviation is -130 µm.
Next you find the tolerance width in table 1. For a 50 mm IT8 the toler-
ance width is 39 µm.
From this you can calculate the tolerance limits:
Upper limit: 50 mm - 130 µm = 49,870 mm
Lower limit: 50 mm - (130+39) µm = 49,831 mm
Tables for the most common tolerance codes can be found in ISO 286-
2. You will find 50 mm c8 in table 17 where the limits are given as -130
µm to -169 µm.

Example: Hole tolerance 20 mm K7

You find the hole tolerance for a 20 mm K7 tolerance in ISO 286-1:2010
as follows:
First you find the fundamental deviation in table 2. For a 20 mm K7
tolerance you first find a fundamental deviation of -2 µm and then a
delta value of +8 µm, so the final fundamental deviation is (-2+8) µm
= +6 µm.
Next you find the tolerance width in table 1. For a 20 mm IT7 the toler-
ance width is 21 µm.
From this you can calculate the tolerance limits:
Upper limit: 20 mm + 6 µm = 20,006 mm
Lower limit: 20 mm + (6-21) µm = 19,985 mm
Tables for the most common tolerance codes can be found in ISO 286-
2. You will find 20 mm K7 in table 8 where the limits are given as +6
µm to -15 µm.

Find your way in GPS 47

 2.9 Former practice
Before ISO 14405-1 was published, the rules for size requirements were sig-
nificantly different. If you work with drawings issued before 2010, you have
to apply these rules.
If the tolerance was indicated with numeric values as shown in figure 2–1 and
2-2 without any further indications, the detailed meaning of the tolerance
was not defined, i.e. the ISO standards did not specify whether the tolerance
should be interpreted as a two-point size, an envelope requirement or some-
thing different.
If the tolerance was indicated with ISO 286-1 tolerance codes, the envelope
requirement applied to the maximum material size of the tolerance, and two-
point size applied to the other side of the tolerance.
A third method was to refer to ISO 8015 in the drawing by writing “Toleran-
cing ISO 8015” close to the title block of the drawing, see figure 2–14.

Figure 2–14:  Indication of ISO 8015 near the title block of the
drawing

In this case the rules for size requirements were identical to the current rules,
i.e. the default was two-point size and the envelope requirement applied when
was indicated after the tolerance, as in figure 2–13.
The “Tolerancing ISO 8015” indication near the title block can still be used,
but is now only informative. The new version of ISO 8015 that was published
in 2011 provides this indication as an optional, informative indication to em-
phasise that the drawing is prepared in accordance with the ISO GPS system.

48 Find your way in GPS

Before ISO 286-1:2010 was issued, ISO 286-1:1988 applied for tolerances
indicated with tolerance codes. It referenced ISO/R 1938:1971, Inspection
of plain workpieces. This is important, because ISO/R 1938 gave manufac-
turing tolerances and permissible wear for the gauges used to measure size
tolerances and this changed the legal tolerance limits for size tolerances. It
extended the tolerance range by over 20 % in many cases.

2.10 Dimensions other than size

Figure 2–15 shows an example of a component that is toleranced with dimen-
sional tolerances.

Figure 2–15:  Component toleranced with dimensional

tolerances

On one hand, the drawing has sufficient dimensions to allow you to calculate
the nominal (theoretical) location of each feature in the component. On the
other hand, there are many problems regarding the tolerancing of this com-
ponent.
The only true features of size are the holes, so the only size tolerances are the
diameter tolerances (¤10±0,05 mm and 4× ¤5±0,05 mm).
You might think that the feature that the 30±0,3 mm tolerance applies to on
the left side of the component is a feature of size, but because the two surfaces
that make up the feature do not have the same extent, it is not a true feature
of size.

Find your way in GPS 49

 The problems with dimensional tolerancing are based on two fundamental
limitations.
The first limitation is that it is impossible to know exactly from where and to
where many of the dimensions go.
Figure 2–16 shows the step width tolerance from figure 2–15 and the problem
in interpreting the tolerance. The ISO standards do not define how to inter-
pret the tolerance. This makes it unclear whether the measurement direction
follows one or the other of the two surfaces and where along the surfaces the
measurement shall be made.

Figure 2–16:  Dimensional tolerances for steps are not well

defined

Figure 2–17 shows the tolerance for the distance from the edge to the hole
from figure 2–15 and the problem interpreting the tolerance. The ISO stan-
dards give no definition for interpretation of this tolerance. Therefore it is
unclear whether it is the distance between median lines (e.g. least-squares)
through the surface and the hole, or if it is the distance between features that
are entirely outside the material, or it is the distance to the nearest point.

Figure 2–17:  Distances are not well defined

50 Find your way in GPS

Figure 2–18 shows the radius tolerance from figure 2–15 and the problem
interpreting the tolerance. The ISO standards give no definition for interpre-
tation of the tolerance. So it is not clear whether it is the smallest radius you
can find, or the largest, or you fit a circle so it covers the entire corner and is
either entirely inside or entirely outside the material.

Figure 2–18:  Radius tolerances are not well defined

The other limitation in dimensional tolerancing is that it does not ensure that
the dimensions are aligned with each other or perpendicular to each other,
although it may look that way in the drawing. Figure 2–19 illustrates this
problem.

All the dimensions can be within the tolerances, but the component can be
out of square.

Find your way in GPS 51

 Figure 2–19:  Dimensional tolerances do not control angles

2.11 Summary
In this chapter we have covered dimensional tolerances. We have defined fea-
tures of size and shown why dimensional tolerances shall only be used for fea-
tures of size, if you want to be entirely sure that your dimensional tolerances
are unambiguous and cannot be misinterpreted.
We have covered the different ways to indicate size tolerances, both with di-
rect indication of numerical limits and by the use of ISO codes.
We have seen that since ISO 14405-1 was published in 2010, a dimensional
tolerance applies to the 2-point dimension by default and that you can change
it to an envelope requirement by indicating an after the tolerance. With

52 Find your way in GPS

the publication of ISO 14405-1 the meaning of a tolerance is now the same
whether you indicate numerical limits or use ISO codes.
We saw that before ISO 14405-1 was published, the meaning of a dimensional
tolerance was not defined in details, unless you either indicated that ISO 8015
applied or used ISO tolerance codes.

Find your way in GPS 53

54 Find your way in GPS
3 Tolerance zones

Geometrical tolerances are based on tolerance zones. We can set up very pre-
cise requirements for a component’s features and their relative location and
orientation by defining tolerance zones inside which the features shall remain
and by attaching these tolerance zones to some of the component’s other fea-
tures.
Let us first look at the three examples from chapter 2, where the dimensional
tolerances were ambiguous and see how we can make the requirements un-
ambiguous with geometrical tolerances.

Figure 3–1:  Step height toleranced with a tolerance zone I

The left picture in figure 3–1 shows one way to tolerance the step height from
chapter 2 with a geometrical tolerance. We have first appointed the long sur-
face as datum D (we will cover datums in chapter 5). Next we have used a
“boxed dimension”, or a TED (theoretically exact dimension) as it is really
called, i.e. a theoretically exact dimension of 15 mm that indicates the nomi-
nal distance between the planes. Finally, we have indicated a position toler-
ance with a width of 0,2 mm for the short surface relative to datum D, i.e. the
long surface.
We can see what it means in the right picture. The leftmost red line is outside
the datum feature, as if the datum feature was put against a perfect counter-
part. The rightmost red line is in the nominal distance from the datum line

Find your way in GPS 55

 (the leftmost red line). The tolerance zone is limited by the two blue lines.
The two blue lines are positioned symmetrically around the nominal position
(the rightmost red line). This tolerance is completely unambiguous: The short
surface shall be contained between the blue lines for the component to pass
the tolerance, so the component in figure 3–1 passes this tolerance.

Figure 3–2:  Step height toleranced with a tolerance zone II

But it could also be the case that our function required the use of the short
surface as the datum and the long surface as the toleranced feature. This is
shown in figure 3–2. As you can see, in this case the toleranced feature falls
entirely outside the blue tolerance zone. Because it is not completely inside
the tolerance zone, the component does not pass this tolerance.
So we can see that with geometrical tolerances, we can indicate our geometri-
cal requirements for the component very precisely and unambiguously. In
most cases we would probably use the tolerance in figure 3–1 rather than
the one in figure 3–2, because generally it is better to use the large surfaces
as datums, because they are more stable. However, it is always the function
that shall dictate how to tolerance, so maybe figure 3–2 better expresses the
functional requirements for the component.

Figure 3–3:  Hole distance toleranced with a tolerance zone

56 Find your way in GPS
The left picture in figure 3–3 shows one way of tolerancing the hole distance
from chapter 2 with a geometrical tolerance. First we have appointed the bot-
tom surface as datum A. Next we have used a TED of 15 mm to indicate the
nominal distance from the bottom surface to the hole. Finally, we have indi-
cated a position tolerance with a width of 0,2 mm for the median line of the
hole relative to datum A, i.e. the bottom surface.
The right picture in fig. 3-4 shows what this means. The upper red line is in
the nominal distance from the datum line (the lower red line). The tolerance
zone is limited by the two blue lines. The two blue lines are positioned sym-
metrically around the nominal position (the upper red line). This tolerance
is completely unambiguous: The median line of the hole shall be contained
between the blue lines for the component to pass the tolerance, so this com-
ponent passes this tolerance.
Since this introduction is only intended to give a first impression of how geo-
metrical tolerances and tolerance zones work, there are a number of details
that we have not covered here. We will get back to those in chapter 4.

Figure 3–4:  Radius toleranced with a tolerance zone

The left picture in figure 3–4 shows a way of tolerancing the radius from chap-
ter 2 with a geometrical tolerance. First we have used a TED of 5 mm to indi-
cate the nominal radius of the surface. Then we have indicated a surface form
tolerance with a width of 2 mm for the surface.
What that means is shown in the right picture. The red line shows the nomi-
nal radius for the surface, and the tolerance zone limited by the two blue lines
is positioned symmetrically in the distance ± t/2 around the nominal radius.
This means that the inner limit has a radius of 4 mm and the outer limit has
a radius of 6 mm. This tolerance is completely unambiguous: The curved sur-
face shall be contained between the blue lines for the component to pass the
tolerance, therefore this component passes this tolerance.

Find your way in GPS 57

 The tolerance zone is not locked by datum references, so it can move about
and adjust to fit the toleranced feature, but it has to maintain its radius. We
could have locked the tolerance zone to datums, if that was what the function
required. We will discuss that in chapter 4.
The purpose of these examples was only to give a small impression of how
geometrical tolerances work and how tolerance zones are used. In the rest of
this chapter we will focus on the different shapes of tolerance zones.
We will examine the details of the different tolerance symbols (we used posi-
tion and form of a surface in these examples) and how to determine the shape
and width of the tolerance zone in chapter 4. We will cover datums and how
they lock tolerance zones in details in chapter 5.
We use the degrees of freedom indicator to show which degrees of freedom
are redundant, locked and unlocked for each tolerance zone. The degrees of
freedom indicator is not a standardised symbol. Annex B covers the meaning
of the degrees of freedom indicator, so you can find an explanation there, if
the meaning is unclear in this chapter. In fact, you might want to read annex
B before continuing with this chapter.

3.1 3D tolerance zones

Most geometrical tolerances are based on 3D tolerance zones, i.e. tolerance
zones that define a volume that shall contain the toleranced feature.
Another possibility is a 2D tolerance zone, i.e. a tolerance zone that defines an
area in a surface that shall contain the toleranced feature.

3.1.1 Two parallel planes

Figure 3–5:  3D tolerance zone for a nominally plane

surface limited by two parallel planes
58 Find your way in GPS
Figure 3–5 shows a very common situation. The toleranced feature is a nomi-
nally flat surface and the tolerance applies to the entire surface. This could for
example be a flatness tolerance, a parallelism tolerance or a position tolerance
for the surface.
In this situation, the tolerance zone is limited by two parallel planes. The dis-
tance between the planes is equal to the tolerance value t. The extent of the
planes is in principle infinite in both directions. The toleranced feature shall
be contained within the volume between the two planes.
Because they are infinite, the planes look the same, whether they move in the
two directions in the plane or rotate around the normal to the plane. There-
fore we say that these three degrees of freedom are redundant, and therefore
they cannot change the tolerance zone.
In chapters 4 and 5 we will look at degrees of freedom for tolerance zones
and how you can use your drawing indication to control which degrees of
freedom are locked and which are free, so the tolerance indications can move.
This way you can control how the tolerance zone locks the toleranced feature
and how it allows the feature to move.
It is important to keep track of which degrees of freedom are redundant for
a tolerance zone, firstly because the tolerance zone can never lock the toler-
anced feature in these directions and secondly because locking the tolerance
zone in a redundant direction does not change the meaning af the tolerance.

Figure 3–6:  3D tolerance zone for a nominally straight

line limited by two parallel planes

In figure 3–6 the toleranced feature is a straight line and the tolerance zone
is limited by two parallel planes. This situation only occurs if the line is a
derived feature, for example the median line in a shaft, and if you have not
written ¤ in front of the tolerance value to indicate that the tolerance zone is
cylindrical. The distance between the planes is equal to the tolerance value t.

Find your way in GPS 59

 It could for example be a straightness tolerance, a perpendicularity tolerance,
or a position tolerance for the median line.
The extent of the planes is again infinite in both directions, so the tolerance
zone has the same redundant degrees of freedom as the one in figure 3–5.
This means that the tolerance does not limit the median line in the directions
perpendicular to the tolerance width.

3.1.2 A cylinder

Figure 3–7:  3D tolerance zone for a nominally straight

line limited by a cylinder

Figure 3–7 shows the situation where the toleranced feature is a straight line
and the tolerance zone is limited by a cylinder. This situation only occurs if
the line is a derived feature, for example the median line in a shaft and if you
have written ¤ in front of the tolerance value to indicate that the tolerance
zone is cylindrical. The diameter of the cylinder is equal to the tolerance value
t. It could for example be a straightness tolerance, a perpendicularity toler-
ance, or a position tolerance for the median line.
Because its length is infinite, the cylinder looks the same, if it moves along its
axis. It also looks the same, if it rotates around its axis. These two degrees of
freedom are therefore redundant.

60 Find your way in GPS

3.1.3 Two coaxial cylinders

Figure 3–8:  3D tolerance zone for a cylindrical feature,

limited by a pair of coaxial cylinders

Figure 3–8 shows the situation where the toleranced feature is a cylindrical
feature and the tolerance zone is limited by a pair of coaxial cylinders. The
difference in radius between the cylinders is equal to the tolerance value t. In
most cases this would be a cylindricity tolerance, but it could also be a total
radial run-out tolerance.
Because their length is infinite, the cylinders look the same, if they move
along their axis. They also look the same, if they rotate around their axis.
These two degrees of freedom are therefore redundant.
However, this type of tolerance zone has an extra degree of freedom, because
the average diameter of the two cylinders can vary and adjust to fit the toler-
anced feature, as long as the radius difference remains equal to the tolerance
value.

Find your way in GPS 61

 3.1.4 A sphere

Figure 3–9:  3D tolerance zone for a point limited by a sphere (the

three copies show the three redundant degrees of freedom)

Figure 3–9 shows the situation where the toleranced feature is a point and the
tolerance zone is limited by a sphere. This situation only occurs, if the point
is a derived feature, for example the centre of a sphere and if you have written
S¤ (for spherical diameter) in front of the tolerance value to indicate that the
tolerance zone is spherical. The diameter of the sphere is equal to the toler-
ance value t. This can only be a position tolerance for the sphere centre.
The sphere looks the same if it rotates around its centre in any direction.
The three rotatory degrees of freedom are therefore redundant. Figure 3–9
shows three copies of the tolerance zone to show the three redundant degrees
of freedom, i.e. the rotations around the three mutually perpendicular axes
through the centre.

3.1.5 Two “parallel” surfaces

In the vast majority of cases the toleranced feature is a plane surface, a cylin-
der, a straight line or a point. However, it is possible to use geometrical toler-
ances for features of any shape. It is necessary to define the nominal geometry
before you can use geometrical tolerances for a feature other than the shapes
mentioned.
Figure 3–10 shows how the tolerance zone is defined in this case: A sphere
with the tolerance value t as its diameter and its centre on the nominal geo-
metry is rolled across the entire nominal geometry to form the tolerance
limits. This definition means that the inner tolerance limit in each arc has a
smaller radius than the outer tolerance limit.

62 Find your way in GPS

 Figure 3–10:  Definition of a 3D tolerance zone for a
free-form surface

Figure 3–11 shows the situation where the toleranced feature is a free-form
surface and the tolerance zone is limited by such a pair of “parallel” surfaces
that surround the toleranced feature.

Form tolerance:
ly li-
A tolerance that on
e fea-
mits the form of th
ienta-
ture, but not its or
tion or location.

This situation can only occur with a form of a surface tolerance. In this exam-
ple the tolerance zone only consists of rounded arcs, but the corner becomes
sharp, if the nominal geometry has corners, where radius is less than half the
tolerance value. Outside corners will always be round.

Find your way in GPS 63

 Derived feature:
that
Theoretical feature
cted)
is derived (constru
tegral
from one or more in
ature
features. A derived fe
form
can be ideal (without
(with
error) or non-ideal
form error).

Figure 3–11:  3D tolerance zone for a free-form surface

limited by two “parallel” surfaces

In the example in figure 3–11, the nominal profile looks like a roof tile, so it
is “alike” in one direction. Therefore it has one redundant degree of freedom.
This is peculiar to this example. Free-form surfaces will have between zero
and three redundant degrees of freedom. Zero, if it is shaped in both direc-
tions; one, if it is straight in one direction; two, if it is a part of a sphere or a
part of a cylinder; and three, if it is a plane surface. The last one is, of course,
a degenerated example, but it shows that the rules are continuous when you
go from one geometry type to another.
64 Find your way in GPS
3.2 Set of 2D tolerance zones
A 3D tolerance zone limits the entire toleranced feature to be within a defined
3 dimensional space limited by a surface (a sphere or a cylinder) or two paral-
lel surfaces.
A set of 2D tolerance zones limits each line in a given direction in the feature
to a 2 dimensional space limited by two lines or a circle.
The individual 2D tolerance zone can move relative to the other zones in the
set. A 2D tolerance is therefore less restrictive than the corresponding 3D
tolerance with the same tolerance value for a feature.
A set of 2D tolerance zones is best used as a supplement to a 3D tolerance
zone. In this case the 2D tolerance shall have a smaller tolerance value than
the 3D tolerance to create an additional requirement. If the two have the same
value, the 2D requirement is automatically fulfilled, if the 3D requirement is
fulfilled and is thus superfluous.
For some 2D tolerances it is possible that they could apply in different direc-
tions, for example in the case of straightness of a flat surface. In this case the
rule is that the requirement applies in the drawing plane for the drawing pic-
ture where the indication is given.

Figure 3–12:  2D tolerance zones apply in the drawing

plane for the drawing picture where the indication is given

3.2.1 Set of parallel, straight line pairs

Figure 3–13 shows a very common situation. The toleranced feature is a nom-
inally plane surface and the tolerance applies to each line element in the plane
surface in a given direction. It could for example be a straightness tolerance
for a plane surface.
In this case each tolerance zone is limited by two parallel, straight lines. The
distance between the lines is equal to the tolerance value t. The lines extend,
in principle, to infinity, so the degree of freedom for the movement along the
lines is redundant.

Find your way in GPS 65

 The tolerance zones are aligned by the projection plane and are perpendicular
to the toleranced feature. This automatically locks three degrees of freedom
for each tolerance zone.

Figure 3–13:  Set of 2D tolerance zones, each of which is

limited by two parallel, straight lines with the distance t

Each line in the toleranced feature shall remain within “its” two lines.
If no datums have been referenced, each tolerance zone can move up and
down and rotate in the intersection plane, independently of the other toler-
ance zones. Therefore they each have two unlocked degrees of freedom.
This type of tolerance only locks the toleranced feature in one direction. It can
be used as a supplement to the tolerance zone shown in figure 3–5, if there are
special requirements for the surface in one direction. It can also be used as a
supplement to the tolerance zone shown in figure 3–11, if the requirement for
the toleranced feature has to be tighter in the “straight” direction.

66 Find your way in GPS

3.2.2 Set of pairs of concentric circles
Figure 3–14 shows a situation that primarily occurs in a roundness toler-
ance or a radial circular run-out tolerance (not total run-out). The toleranced
feature is cylindrical (hole or shaft) and the tolerance applies to each cross-
section of the cylinder.
Each tolerance zone is limited by two concentric circles. The difference in
radius between the two circles is equal to the tolerance value t. Note that it is
not the diameter difference. The 2-point diameter in the cross-section of the
toleranced feature can vary by twice the tolerance value.
The tolerance zones are aligned by the cylinder axis. This automatically locks
three degrees of freedom for each individual tolerance zone. The rotation
around the axis is redundant, because the circles look the same when you
rotate them. So unless they are locked by datums (which is not the case with
roundness), each tolerance zone can move in the two directions in the inter-
section plane. In addition, the average diameter for the two circles can vary.
This gives an extra degree of freedom for this type of tolerance zone.

of
Non-redundant degree
freedom:
r a tole-
Degree of freedom fo
change in
rance zone where a
ee of free-
the value of the degr
tolerance
dom changes the
zone.

Find your way in GPS 67

 Figure 3–14:  Set of 2D tolerance zones, each of which is limited
by two concentric circles with a radius difference of t

68 Find your way in GPS

3.2.3 Set of “parallel” non-straight pairs of lines
Figure 3–15 shows the last case for a set of 2D tolerance zones. The toleranced
feature is neither a cylinder, nor a sphere, nor a nominally plane surface and
the tolerance applies to each line element in the surface in a given direction.
It is necessary to define the nominal geometry, if the feature is not one of
the mentioned geometries before you can use geometrical tolerances for the
feature.
A circle with the tolerance value t as its diameter and its centre on the nomi-
nal geometry is rolled along the entire nominal profile in each cross-section
to form the tolerance limits. This is similar to how a sphere was used in figure
3–10. As in figure 3–10 this definition means that the inner tolerance limit in
each arc has a smaller radius than the outer tolerance limit.
The tolerance zones are aligned by the projection plane and are perpendicular
to the toleranced surface. This automatically locks three degrees of freedom
for each tolerance zone. The tolerance zone has no redundant degrees of free-
dom in the given case. This is typical for free-form profile tolerances.

freedom:
Redundant degree of
for a tolerance
A degree of freedom
in the value of
zone where a change
m does not chan-
the degree of freedo
ne. For example,
ge the tolerance zo
e zone has two
a cylindrical toleranc
freedom; move-
redundant degrees of
of the tolerance
ment along the axis
ound the axis.
zone and rotation ar

Find your way in GPS 69

 Figure 3–15:  Set of 2D tolerance zones, each of which
is limited by two “parallel”, non-straight lines with the
distance t

3.3 Individual 2D tolerance zones

Most individual 2D tolerance zones are used to tolerance either a linear fea-
ture or a specific line in a surface feature. In this case the 2D tolerance zone
limits the toleranced line to a 2-dimensional space that is limited by two lines.
The other possibility is that the 2D tolerance zone applies to a point on a
linear feature, for example a point on the median line of a cylinder. In this

70 Find your way in GPS

case the 2D tolerance zone limits the toleranced point to a 2-dimensional
space that is limited by a circle, or in rare cases, two lines.

One pair of parallel straight lines

Figure 3–16 shows the situation where the toleranced feature is a generatrix
in a cylindrical surface, and the tolerance only applies to a specific generatrix.
In this case the tolerance zone is limited by two parallel lines. The distance
between the lines is equal to the tolerance value t. The lines are in principle
infinitely long, so the degree of freedom for movement along the lines is re-
dundant.
The tolerance zone is aligned by the intersection plane containing the cylin-
der axis and is perpendicular to the toleranced surface. This automatically
locks three degrees of freedom for the tolerance zone.

Figure 3–16:  2D tolerance zone limited by two parallel,

straight lines with the distance t

If no datums are referenced, the tolerance zone can move up and down and
rotate in the intersection plane. It therefore has two unlocked degrees of free-
dom.

A circle
Figure 3–17 shows a situation where the toleranced feature is a point on the
median line of a cylinder, and the tolerance only applies to this point.
This type of tolerance could for example be used to ensure that the centre of a
hole in a tube is concentric with the outer diameter in a given cross-section. If

Find your way in GPS 71

 you make such a requirement for the cross-section of the tube, you can ensure
that the wall thickness is uniform around the tube.
In this case, the tolerance zone is limited by a circle. The diameter of the circle
is equal to the tolerance value t. The circle does not change if you rotate it, so
the degree of freedom for the rotation is redundant.
The tolerance zone is aligned by the intersection plane, which is perpendicu-
lar to the cylinder axis. This automatically locks three degrees of freedom for
the tolerance zone.

Figure 3–17:  2D tolerance zone limited by a circle with

diameter t

Without datums, the circle can freely move in the intersection plane and does
not limit the toleranced feature (the point on the median line) at all. So it is
necessary to use datums for this type of tolerance zone in order to be mean-
ingful. If no datums are referenced, the tolerance zone can move in the two
directions in the intersection plane. Therefore it has two unlocked degrees of
freedom. In chapter 4 we will look at a concentricity example using this type
of tolerance zone.

A pair of concentric circles

Figure 3–18 shows a situation where the toleranced feature is a particular
cross-section of a cylinder and the tolerance only applies to this cross-section.
In this case the intersection plane and therefore the position of the toleranced
cross-section has to be defined. This is typically done by referencing datums.

72 Find your way in GPS

The tolerance zone is aligned by the cylinder axis. This automatically locks
three degrees of freedom for the tolerance zone. The rotation around the axis
is redundant, because the circles look the same when you rotate them. So
unless they are locked by datums (which is not the case for roundness), each
tolerance zone can move in the two directions in the intersection plane. In
addition, the average diameter of the two circles can vary. This gives an extra
degree of freedom for this type of tolerance zone.
This situation is quite rare, but is included for completeness.

Figure 3–18:  2D tolerance zone limited by two concentric

circles with t as their radius difference

One pair of “parallel” (non-straight) lines

Figure 3–19 shows the last case for an individual 2D tolerance zone. The tol-
eranced feature is in this case an identified line in a free-form surface and the
tolerance only applies to this line element. Typically you have to identify the
line element with theoretically exact dimensions (TEDs) referencing datums.
Likewise, the nominal geometry of the line has to be defined before you can
use geometrical tolerances for the feature.
In the same way as in figure 3–15, you use a circle with the tolerance value t
as its diameter and its centre on the nominal geometry and roll it along the
entire nominal profile in the identified cross-section to form the tolerance
limits.
As in figure 3–15 this definition means that the inner tolerance limit in each
arc has a smaller radius than the outer tolerance limit.

Find your way in GPS 73

 The tolerance zone is aligned by the intersection plane that defines the toler-
anced line and is perpendicular to the toleranced surface. This automatically
locks three degrees of freedom for the tolerance zone. The tolerance zone in
the example has no redundant degrees of freedom. This is typical for free-
form profile tolerances.

Figure 3–19:  2D tolerance zone limited by two “parallel”,

non-straight lines with a distance of t

3.4 Summary
In this chapter we have covered the various types of geometrical tolerance
zones. We have seen that these tolerance zones fall in three categories:
"" 3D tolerance zones that are typically limited by two surfaces and lock
a toleranced feature to remain between the two surfaces. Other pos-
sibilities are a cylindrical tolerance zone that applies to a line, and a
spherical tolerance zone that applies to a point.

"" Infinite sets of 2D tolerance zones, limited by two lines locking each
their line in a surface

"" Individual 2D tolerance zones that either lock an identified line in a

surface or a point on a line.

74 Find your way in GPS

 Integral feature:
forms
Real feature that
of the
part of the surface
component.

We have seen that the toleranced feature can be a surface or a line in a surface
on the component. We call these features integral features. The toleranced fea-
ture can also be a median surface, a median line in a cylinder or a point on a
median line. We call these features that are not on a surface on the component
derived features.
We now have an overview of the various types of tolerance zones. In the next
chapter we will cover the tolerance symbols used to define tolerance zones.
We will also look at how you can either lock the tolerance zones to other fea-
tures on the component, or let them move and adjust to fit the toleranced fea-
ture, depending on what the functional requirements for the component are.

Find your way in GPS 75

76 Find your way in GPS
4 Tolerance symbols

GPS tolerances are based on tolerance zones. In the last chapter we looked at
the different types of tolerance zones.
Information about the type, width, locked degrees of freedom of the tolerance
zone is given in the tolerance indicator.
In this chapter we will cover the 14 tolerance symbols used for geometrical
tolerances in GPS.
The tolerance symbol is the first thing we look at to get information about the
tolerance type, when we read a drawing and start getting an overview of the
geometrical tolerances. The tolerance symbol is indicated in the first box in
the tolerance indicator.
Designers use a combination of the different types of geometrical tolerances
to ensure that the various features of the component are in their correct loca-
tion and orientation and have sufficiently small form deviations to ensure the
function of the product.

Figure 4–1:  The tolerance indicator

The position symbol in the first box tells us that the tolerance indicator in
figure 4–1 specifies a position tolerance.
In the second box it is indicated that the tolerance zone is cylindrical, because
there is an ¤ in front of the tolerance value, which therefore is the diameter of

Find your way in GPS 77

 the tolerance zone. The M indication tells us that the tolerance is modified
with a maximum material requirement (we will cover that in chapter 7).
The third box tells us that datum A is the primary datum and because of the
M indication here, datum A is part of the maximum material requirement.

The fourth box indicates that the secondary datum is a common datum,
which is based on datums C and B (we cover datums and what they mean in
chapter 5).
Finally, we can see in the fifth box that datum K is the tertiary (third) datum.
In this chapter we will primarily focus on the symbol in the first box and the
information it gives us.
GPS tolerances can be divided into:
•• Unrelated form tolerances not referring to datums or datum systems
•• Related tolerances referring to datums or datum systems. Related tol-
erances can further be divided into:
−− Mobile tolerances, or orientation tolerances, that have some, but
not all degrees of freedom locked to datums or datum systems
−− Fixed tolerances, or location tolerances, that have all degrees of
freedom locked to datums or datum systems
•• Run-out tolerances combining form and location.

78 Find your way in GPS

 Degree of freedom:
at de sc rib es th e po sition of a compo-
Parameter th at e of a tolerance zone
.
e po si ti on or st
nent or th dom.
je ct s ha ve si x po si tional degrees of free
All ob coor-
e de sc rib e th e lo ca tion of the object as
Thes
Ca rt es ia n co or di na te system and its
dinates in a ound the three axes
in
io n as ro ta ti on s ar
orientat nce
ordi na te sy st em . In addition, some tolera
the co that allow
ne s ha ve in tr in si c degrees of freedom r
zo
ju st to fit th e to leranced feature. Fo
them to ad er of a cylindricity to
-
th e m ed ia n di am et
example,
lerance.
an sl at or y de gr ee s of freedom describe
The three tr coordinates in a Cart
e-
ti on of th e ob je ct as
the loca grees
co or di na te sy st em , and the rotatory de
sian ion as ro-
ee do m de sc rib e the object’s orientat
of fr
nd th e th re e ax es in the coordinate sy-
tations arou
stem.

Find your way in GPS 79

 Mobile related toler-

Unrelated tolerance

Related run-out toler-

Fixed related tolerance

Surface requirement
Line requirement

Derived features
Without datums
ance orientation

Integral features
With datums
location

form

ance
Position √ (√)1 √ √ (√)2 √ √ √

Surface profile √ √ √ √ √ √ √ √

Line profile √ √ √ √ √ √ √ √

Coaxiality √ √ √ √

Symmetry √ √ √ √ √ √

Parallelism √ √ √ √ √ √

Perpendicularity √ √ √ √ √ √

Angularity √ √ √ √ √ √

Flatness √ √ √ √ √

Cylindricity √ √ √ √ (√)3

Straightness √ √ √ √ √

Roundness √ √ √ √ (√)3

Total run-out √ √ √ √

Circular run-out √ √ √ √

1: The position symbol can create a mobile related tolerance, if all of the degrees of freedom for
the tolerance zone are not locked by the referenced datums.

2: The position symbol can be used without reference to datums in TED patterns.

3: In principle, you can tolerance e.g. the roundness of the major circle in a torus or the round-
ness and/or the cylindricity of the median cylinder in a tube, but apart from such very rare
situations roundness and cylindricity are only used for integral features.
Table 4–1:  Tolerance symbols
80 Find your way in GPS
In the following we will look at a number of examples of the different toler-
ance symbols to illustrate which types of tolerance zone (or set of tolerance
zones) they can each indicate.
Most of the symbols require reference to one or more datums to be meaning-
ful. In this chapter we will not go into details about how datums are created
from the datum features. We will instead cover this in depth in chapter 5.
The number and type of the referenced datums change how the tolerance
zone can adjust to fit the toleranced feature. We will also save this topic for
chapter 5.
When tolerancing, normally you will start with fixed tolerances that lock the
location of the component’s features. Fixed tolerances can be used alone, if
there are no additional requirements for the orientation and form of the fea-
ture.
If the fixed tolerances are not sufficient to ensure the function of the compo-
nent, you will supplement them with mobile tolerances (typically orientation
tolerances) and unrelated form tolerances.
Finally, in some cases you would combine the form and location tolerances
into run-out tolerances.
We will use this sequence to cover the tolerance symbols in this chapter. Be-
cause some of the symbols (position, line profile and surface profile) can be
used for several types of tolerances, examples of where they can be used are
given in each section.
We use the degrees of freedom indicator to show which degrees of freedom
are redundant, locked and unlocked for each tolerance zone, as mentioned
in the beginning of chapter 3. The degrees of freedom indicator is not a stan-
dardised symbol. Annex B covers the meaning of the degrees of freedom in-
dicator, so if the meaning is unclear in this chapter you can find an explana-
tion there. In fact, if you have not done so already, you may want to read
annex B before continuing with this chapter.

Find your way in GPS 81

 Datum feature:
which
Real feature from
d
the datum is derive

4.1 Fixed tolerance zones

Fixed tolerance zones are tolerance zones, where the referenced datums lock
all non-redundant degrees of freedom. The position, surface profile, line pro-
file, coaxiality and symmetry symbols can be used to create fixed tolerance
zones that lock the location and orientation of the feature.
Run-out tolerance is a special type of fixed tolerances, which will be covered
separately later in this chapter.

Translatory degree of
freedom:
A degree of freedom that
allows linear motion in one
direction. As opposed to
a rotatory degree of free-
dom

82 Find your way in GPS

Position

Position is a fixed related tolerance when the referenced datums lock all non-
redundant degrees of freedom for the tolerance zone.
A fixed related tolerance locks the location and orientation of the toleranced
feature relative to one or more datums.
Position tolerances can create surface requirements or line requirements. Po-
sition tolerances can be used for both integral features and derived features.
Position tolerances are the primary tool we use for locking features in GPS
tolerancing.

Position example 1: Position of a flat surface

Figure 4–2 shows an example of a position tolerance for a real flat sur-
face (an integral feature). The position tolerance applies to the entire
flat surface.

Figure 4–2:  Position tolerance for a flat surface related to

a datum

Find your way in GPS 83

 Figure 4–3:  Explanation drawing for figure 4–2

The position tolerance controls both the location (distance), orienta-

tion (parallelism) and form (flatness) for the toleranced feature and can
therefore be used on its own.

84 Find your way in GPS

Position example 2: Position of the median line of a hole
Figure 4–4 shows an example of a position tolerance for the median
line of a hole (a derived, non-ideal feature). The tolerance controls the
location, orientation and straightness of the median line, but not the
geometry of the hole itself. The tolerance references a datum system,
which consists of two datums that lock the tolerance zone completely.

Figure 4–4:  Position tolerance for the median line of a

hole related to a datum system, cylindrical tolerance zone

Rotational degree of
freedom:
that
A degree of freedom
ound
allows rotation ar
d to
an axis. As oppose
ee of
a translatory degr
freedom.

Find your way in GPS 85

 Figure 4–5:  Explanation drawing for figure 4–4

Position tolerances like the one in figure 4–4 are used to completely
lock the location and orientation of the median line relative to primary
datum B and secondary datum A.

Related tolerance:
feren-
A tolerance that re
ces datums.

86 Find your way in GPS

Position example 3: TED pattern consisting of 4 holes with a
datum system consisting of 3 datums
Figure 4–6 shows the component used in this and later examples.

Figure 4–6:  The component used in position example 3

A TED pattern is a pattern of theoretically exact dimensions (TEDs),

which indicate the nominal distances between a number of features.
The drawing in figure 4–8 indicates that the pattern consists of 4 holes
with a horizontal distance of 25 mm and a vertical distance of 20 mm.
Additionally, the drawing shows that the pattern starts 15 mm from the
bottom surface and 15 mm from the right side surface.
You show that the tolerance applies to all 4 holes by writing 4× above
the tolerance indicator.
The referenced datum system ensures 1) that the tolerance zones for the
median lines of the holes are perpendicular to the back side of the plate,
which is indicated as primary datum S, 2) that the pattern is oriented
parallel to secondary datum M in the correct distance and 3) that the
pattern is in the correct distance from tertiary datum P.
This requirement ensures that if the component has to be assembled
with another component like the one in figure 4–7, then the two plates
will be able to contact each other, when they are assembled; the side
surfaces will be parallel and line up; and the top and bottom surfaces
will line up, if both components are toleranced as in figure 4–8, so the
holes and pins are perpendicular to the surface, aligned with the side
surface and in the correct distance from the side surface and bottom
surface.

Find your way in GPS 87

 Figure 4–7:  Counterpart to the component in figure 4–6

Note that the necessary tolerances to the outside surfaces are left out
in the example. Note also that the tolerances only apply to the median
lines and not to the surface of the holes. In chapter 6 we will look at
how the tolerance can be modified to apply to the surface and ensure
assembly.

Figure 4–8:  Position tolerance for a TED pattern, which

consists of the median lines of 4 holes referencing a datum
system consisting of 3 datums

88 Find your way in GPS

Figure 4–9:  Explanation drawing for figure 4–8

Find your way in GPS 89

 Surface profile

Surface profile, or “profile any surface” as ISO 1101 calls it, is a very flexible
tolerance type. You can use it without reference to datums to create a form
requirement for a surface that is neither plane nor cylindrical, as long as the
nominal shape of the surface is defined, e.g. with TEDs.
You can use surface profile for both real (integral) surfaces and derived fea-
tures, such as median surfaces.
You can also use surface profile with reference to datums. This way you can
indicate mobile related tolerances or fixed related tolerances.

Surface profile example 1: Fixed location tolerance for an inte-

gral surface feature
The tolerance in figure 4–10 is an example of a fixed location tolerance,
where the referenced datums lock all non-redundant degrees of free-
dom for the tolerance zone.
The toleranced feature is the periphery of the cam. It consists of two
plane surfaces and two partial cylinders.
According to the principle of independency, these four shall usually
be considered as individual features. Therefore we have to use the “all
around” symbol to indicate that the tolerance applies to the entire pe-
riphery and that it shall be considered as one continuous feature. Had
the tolerance zone been mobile, we would also have had to use “CZ”
common zone indication to show that the tolerance zones for the four
features have to move as one. We will get back to this in clause 4.2.

Ideal feature:
Theoretical feature
that does not have
form error

90 Find your way in GPS

 Figure 4–10:  Surface profile tolerance applied to a
periphery consisting of 4 features with reference to
a datum system that consists of 2 datums

Figure 4–11 illustrates how a surface profile tolerance zone is created

by taking a sphere with the tolerance value as its diameter and roll its
centre along the nominal surface profile. In this case it means that the
inner tolerance limit has a smaller radius than the outer tolerance limit.
The tolerance zone only has soft contours in this example, but the tol-
erance limit on the inside of a radius becomes sharp, if the nominal
geometry has sharp corners, where radius is less than half the tolerance
value. The tolerance limit on the outside of a radius is always rounded.

Line requirement:
Requirement that
applies to a
line or a set of lines
individually.
Straightness an
d roundness
are examples of
line require-
ments. Line requ
irements are
also called 2D
requirements
(two dimension
al require-
ments).

Find your way in GPS 91

 Figure 4–11:  Explanation drawing for figure 4–10

With the tolerance in figure 4–10 we ensure that the periphery of the
cam is located and oriented correctly relative to the hole axis that is
primary datum E and rotated correctly, so it is symmetrical around the
median plane of the keyway, which is datum B.
We would use this tolerance, if the cam was to be mounted on a shaft
and be held in angular position by a key. In this case the tolerance would
ensure that the periphery has the right shape; that it is located correctly
relative to the shaft; and that it is in the correct angular position when
the key is inserted.

Non-ideal feature:
Theoretical or real fea-
ture that has form error.

92 Find your way in GPS

Line profile

Line profile has the same characteristics as surface profile. The difference is
that line profile creates a requirement for each line in the surface individually,
whereas surface profile creates one requirement for the entire surface.

Line profile example 1: Fixed location tolerance for an integral

surface feature
The tolerance in figure 4–12 shows a line profile tolerance for the pe-
riphery of the same cam that was also shown in figure 4–10. The only
difference between the two drawings is that we have used the line pro-
file symbol in figure 4–12 and the surface profile symbol in figure 4–10.

Figure 4–12:  Line profile tolerance applied to a periphery

consisting of 4 features with reference to a datum system
that consists of 2 datums

Figure 4–13 illustrates how a line profile tolerance zone is created by

taking a circle (as opposed to the sphere used for the surface profile tol-
erance zone) with the tolerance values as its diameter and roll its centre
along the nominal line profile. In this case it means that the inner toler-
ance limit has a smaller radius than the outer tolerance limit.
The tolerance zone only has soft contours in this example, but the tol-
erance limit on the inside of a radius becomes sharp, if the nominal

Find your way in GPS 93

 geometry has sharp corners, where radius is less than half the toler-
ance value. The tolerance limit on the outside of the radius is always
rounded.
We can see from the explanation figures 4–11 and 4–13 that the mean-
ings of the two tolerances are different from a formal point of view,
because the surface profile tolerance creates a surface tolerance and the
line profile tolerance creates a set of line tolerances.
But if we study figure 4–13 closer, we can see that all the tolerance zones
for the individual lines are completely locked to the datum system, so
the individual tolerance zone cannot move relative to the other toler-
ance zones. All the line tolerance zones will be perfectly aligned relative
to the datum system and each other. This is why the meaning of the two
tolerances is exactly the same.

Figure 4–13:  Explanation drawing for figure 4–12

94 Find your way in GPS

 The conclusion of this example is that for fixed tolerances it does not
matter whether we use line tolerances or surface tolerances for a real
surface. When there are no unlocked degrees of freedom, the line toler-
ance in reality becomes a surface tolerance.
Later in this chapter, when we look at mobile tolerances and form tolerances,
we will see the difference in meaning between a line tolerance and a surface
tolerance for a real surface, when the tolerance has unlocked degrees of free-
dom.

Coaxiality and concentricity

Coaxiality is a fixed related tolerance that locks all the non-redundant degrees
of freedom for the tolerance zone.
Coaxiality is a specialised position tolerance. It is used exclusively for toleran-
cing median lines (derived features) relative to datum axes.
We can also indicate a concentricity tolerance with the same symbol.

Coaxiality example: Fixed location tolerance for a derived line

feature
The tolerance in figure 4–14 is a typical example of a coaxiality toler-
ance. The tolerance would have had exactly the same meaning, had we
used the position symbol, but by using the coaxiality symbol, we tell the
reader exactly what we want to achieve with our tolerance.

Find your way in GPS 95

 Figure 4–14:  Coaxiality tolerance for the median line in
a cylinder with reference to a common datum, which is a
straight line

Figure 4–15:  Explanation drawing for figure 4–14

96 Find your way in GPS

Concentricity example: Fixed location tolerance for each point
on a derived line feature
As we saw in the surface profile and line profile examples, the meaning
of a tolerance for each line in a surface is the same as a tolerance for the
entire surface, when the referenced datums are the same and there are
no unlocked degrees of freedom.
In the same way a concentricity tolerance will mean the same as a co-
axiality tolerance, if the referenced datums are the same, because there
are no unlocked degrees of freedom for these types of tolerances.
So for the concentricity tolerance to have a different meaning, the da-
tum has to change for each cross-section.
Figure 4–16 shows the classic example of a concentricity tolerance. The
tolerance ensures that the inside of the tube is concentric with the out-
side in each cross-section, regardless of whether the tube is straight or
bent.
This requirement could, for example, be used to ensure that the hole in
a garden hose or an exhaust pipe is centred.
In chapter 7 where we look at least material requirements, we will see
how we can ensure the wall thickness of the garden hose or exhaust
pipe. This not only requires that the hole is centred, but also that it has
the right size and roundness.

Find your way in GPS 97

 Figure 4–16:  Concentricity tolerance for the centre point
of a tube with reference to a datum that is a point

Figure 4–17:  Explanation drawing for figure 4–16

98 Find your way in GPS

Symmetry

Symmetry is a fixed related tolerance that locks all non-redundant degrees of
freedom for the tolerance zone.
Symmetry, like coaxiality and concentricity, is a specialised position toler-
ance. We use symmetry to indicate that the tolerance feature shall be “in the
middle”.

Symmetry example: Fixed location tolerance for a median sur-

face
The tolerance in figure 4–18 is a typical example of a symmetry toler-
ance. The tolerance would have had exactly the same meaning, had we
used the position symbol, but by using the symmetry symbol, we tell
the reader exactly what we want to achieve with our tolerance.
When we use symmetry, we emphasise that there is an unwritten TED
of 0 mm specifying the nominal distance between the datum plane and
the toleranced median surface, i.e. they are coinciding.

Figure 4–18:  Symmetry tolerance for a median surface

with reference to a coinciding datum plane

Find your way in GPS 99

 -
-

Figure 4–19:  Explanation drawing for figure 4–18

4.2 Mobile tolerance zones

Mobile tolerance zones are tolerance zones where the referenced datums lock
some, but not all, non-redundant degrees of freedom.
You can create a mobile tolerance zone in two ways:
One way is to use the orientation symbols, parallelism, perpendicularity and
angularity. These symbols, by definition, only lock the rotational degrees of
freedom, regardless of which degrees of freedom the referenced datums could
lock. This means that by definition such a tolerance zone has at least one un-
locked translatory degree of freedom.
The other way is to use the position, surface profile and line profile symbols,
but not reference enough datums to lock all non-redundant degrees of free-
dom.

100 Find your way in GPS

Which unlocked degrees of freedom such a tolerance has depends entirely on
the referenced datums.

Parallelism

Parallelism is a mobile related tolerance that locks the orientation, but not the
location of the toleranced feature relative to one or more datums. Parallelism
tolerances can create surface requirements or line requirements. Parallelism
tolerances are used for both integral features and derived features.
We typically use parallelism tolerances as a supplement to position tolerances
or size tolerances to ensure better component geometry than these tolerances
do on their own. Parallelism requirements cannot stand alone.

Translatory degree of
freedom:
that
A degree of freedom
in one
allows linear motion
d to
direction. As oppose
free-
a rotatory degree of
dom.

Parallelism example 1: Parallelism of a plane surface

Figure 4–20 shows an example of a parallelism tolerance for a real plane
surface (an integral feature). The parallelism tolerance applies to the
entire plane surface.

Figure 4–20:  Parallelism tolerance for a plane surface

Find your way in GPS 101
 If you compare figure 4–21 to figure 4–3, you can see that the difference
between a parallelism tolerance for a plane with another plane as the
datum and a position tolerance for the same plane with the same datum
is that where the position tolerance zone is completely locked, the par-
allelism tolerance zone has one unlocked degree of freedom.
This means that with the same tolerance value, the position tolerance is
more restrictive than the parallelism tolerance. The position tolerance
ensures parallelism, but the parallelism tolerance does not lock the lo-
cation (the distance to the datum plane).
We use parallelism tolerances as a supplement to position tolerances
when we want the parallelism to be better than what the position toler-
ance ensures.
In other words, we have to use a smaller tolerance for the parallelism
than we used for the position, in order for the parallelism requirement
to be meaningful and add a requirement in addition to the one already
created by the position tolerance.

Figure 4–21:  Explanation drawing for figure 4–20

102 Find your way in GPS

A parallelism tolerance like the one in figure 4–20 is used to lock the
orientation of the toleranced surface.
We will typically use it as a supplement to the position tolerance in
figure 4–2, for example in a case where the datum feature is mounted
against another component in the product and it is important that the
toleranced surface maintains the correct orientation to a tighter toler-
ance than ensured by the position tolerance relative to the rest of the
product after assembly.

Parallelism example 2: Parallelism of the median line of a hole I

Figure 4–22 shows an example of a parallelism tolerance for the median
line of a hole (a derived, non-ideal feature). The tolerance does not con-
trol the geometry of the hole, only the orientation and straightness of
the median line in one direction.

Figure 4–22:  Parallelism tolerance for the median line of

a hole relative to a datum, tolerance zone limited by two
planes

Find your way in GPS 103

 Surface requirement:
applies to a
A requirement that
one or more
surface, as opposed to
lindricity are
lines. Flatness and cy
quirements.
examples of surface re
ts are also
Surface requiremen
ents (three-
called 3D requirem
ents).
dimensional requirem

Figure 4–23:  Explanation drawing for figure 4–22

A parallelism tolerance as the one in figure 4–22 is used to lock the ori-
entation of the toleranced median line, but only in one direction.

104 Find your way in GPS

Note that the tolerance does not lock the parallelism of the median line
to the bottom surface. Even if the hole was drilled vertically through
the component instead of sideways, it would still be able to pass this
tolerance.
The tolerance does not limit the curvature of the median line in the
plane of the tolerance zone, so a hole starting on the top surface, curv-
ing around and exiting through one of the side surfaces could also pass
this tolerance.
So the tolerance zone is not very restrictive, even though it only has one
unlocked degree of freedom, the distance to datum B.
You can only use this tolerance together with another tolerance, for ex-
ample the position tolerance in figure 4–4 that locks the location and
the general orientation of the median line of the hole. In this case, the
parallelism tolerance shall have a smaller tolerance value than the posi-
tion tolerance.
You can use such a tolerance combination, if the position tolerance en-
sures that the median line is sufficiently straight, but it is necessary that
the parallelism of the median line to the datum plane is better than
what the position tolerance requires.
This could be the case, if the component is mounted with the datum
feature against the rest of the product and if it is possible to make ad-
justments, or the component can rotate in the non-toleranced direc-
tion.

Parallelism example 3: Parallelism of the median line of a hole

II
Figure 4–24 shows an example of a parallelism tolerance for the median
line of a hole (a derived, non-ideal feature). The tolerance does not con-
trol the geometry of the hole, only the orientation and straightness of
the median line. The difference from the last example is that the toler-
ance zone is cylindrical.

Find your way in GPS 105

 Figure 4–24:  Parallelism tolerance for the median line of a
hole relative to a datum, cylindrical tolerance zone

The only difference between figure 4–22 and figure 4–24 is the ¤ in
front of the tolerance value in figure 4–24. Figure 4–23 and figure 4–25
show that this small change causes a big change in the meaning of the
tolerance.

106 Find your way in GPS

 Figure 4–25:  Explanation drawing for figure 4–24

A parallelism tolerance like the one in figure 4–24 is used to lock the
orientation of the toleranced median line, but only in one direction.
This tolerance does not lock the parallelism of the median line to the
bottom surface, just like the tolerance in figure 4–22. Even if the hole
was drilled down into the component vertically, instead of horizontally
from the side, it would pass this tolerance. On the other hand, this tol-
erance limits the straightness of the median line in contrast to the toler-
ance in figure 4–22.
Even though this tolerance zone has three unlocked degrees of freedom
instead of only one as the tolerance in figure 4–22 has, it is more restric-
tive.
Furthermore this tolerance can only be used in connection with an-
other tolerance, for example the position tolerance in figure 4–4 that
controls the location and the general orientation of the median line of
the hole. In this case, the parallelism tolerance must have a smaller tol-

Find your way in GPS 107

 erance value than the position tolerance, in order for it to add a require-
ment.
You can use such a tolerance combination, if it is necessary that the
median line is straighter and more parallel with the datum plane than
what the position tolerance ensures on its own.
This could be the case, if the component is mounted with the datum
feature against the rest of the product and it can be adjusted or rotated
in the non-toleranced direction, so it is more important that the me-
dian line is straight and parallel with the datum feature, than it is in the
exact distance from the datum feature.

Perpendicularity

Perpendicularity is a mobile related tolerance that locks the orientation of
the toleranced feature relative to one or more datums. Perpendicularity can
create surface requirements and line requirements and can be used for both
integral features and derived features.
We typically use perpendicularity tolerances as a supplement to position tol-
erances to ensure better component geometry than what these tolerances do
on their own.
We also often use perpendicularity tolerances to ensure the relative orienta-
tion among the features that make up the global datum system in the com-
ponent.

Perpendicularity example 1: Perpendicularity of a plane sur-

face
Figure 4–26 shows an example of a perpendicularity tolerance for a real
plane surface (an integral feature).
The perpendicularity tolerance applies to the entire plane surface.

108 Find your way in GPS

 Figure 4–26:  Perpendicularity tolerance for a plane
surface relative to a datum

The tolerance zone can adjust to

the real surface by moving up or
down and rotating around the
normal to datum D, as long as it
remains perpendicular to datum D

Figure 4–27:  Explanation drawing for figure 4–26

If you compare figure 4–27 to figure 4–21, you will see that the differ-
ence between a perpendicularity tolerance with a plane as the datum
and a parallelism tolerance with a plane as the datum is that where they
both have one unlocked translatory degree of freedom, the perpendicu-
larity tolerance also has an unlocked rotatory degree of freedom. This
means that with the same tolerance value, the parallelism tolerance is
more restrictive.

Find your way in GPS 109

 Note however that because the two tolerances use different datums, the
alignment of the tolerance zones will be different, so they cannot be
compared directly.

Perpendicularity example 2: Perpendicularity of the median

line of a hole I
Figure 4–28 shows an example of a perpendicularity tolerance for the
median line of a hole (a derived non-ideal feature).
The tolerance does not control the geometry of the hole, only the orien-
tation and straightness of the median line in one direction.

Figure 4–28:  Perpendicularity tolerance for the median

line of a hole relative to a datum, tolerance zone limited by
two planes

110 Find your way in GPS

 Figure 4–29:  Explanation drawing for figure 4–28

A perpendicularity tolerance like the one in figure 4–28 is used to lock

the orientation of the toleranced median line, but only in one direction.
By comparing figure 4–23 and figure 4–29, we can see that the toler-
ances in figure 4–22 and figure 4–28 in principle put the same con-
straints on the toleranced median line and have the same limitations as
described above.
However, in reality the two tolerance zones will not be identical, since
the two tolerances use different datum features, and the two real datum
features (B in figure 4–23 and C in figure 4–29) will never be perfectly
perpendicular to each other.

Perpendicularity example 3: Perpendicularity of the median

line of a hole II
Figure 4–30 shows an example of a perpendicularity tolerance for the
median line of a hole (a derived non-ideal feature).

Find your way in GPS 111

 The tolerance does not control the geometry of the hole, only the orien-
tation and straightness of the median line. The difference from the last
example is that the tolerance zone is cylindrical.

Figure 4–30:  Perpendicularity tolerance for the median

line of a hole relative to a datum, cylindrical tolerance zone

The only difference between figure 4–28 and figure 4–30 is an ¤ in front
of the tolerance value in figure 4–30. Figure 4–29 and figure 4–31 show
that this small change causes a big change in the meaning of the toler-
ance.

112 Find your way in GPS

 Figure 4–31:  Explanation drawing for figure 4–30

A parallelism tolerance like the one in figure 4–30 is used to lock the
orientation of the toleranced median line completely.
The tolerance also controls the straightness of the median line, but not
its location. The tolerance is more restrictive than the one in figure 4–28
even though the tolerance zone has two unlocked degrees of freedom
instead of one. This tolerance shall still be used in combination with
another tolerance, for example a position tolerance, to be useful.
Another possibility is the case where the component is mounted with
datum F against another component in the product and a bolt goes
through the toleranced hole and controls the location of the compo-
nent. In this case you would use datum F as the primary datum and the
axis of the hole as the secondary datum in the global datum system for
the component and use the perpendicularity tolerance to ensure that
the secondary datum feature is perpendicular to the primary datum.

Find your way in GPS 113

 Angularity

Angularity is a mobile related tolerance that locks the orientation of the toler-
anced feature relative to one or more datums. Angularity can create surface
requirements and line requirements and can be used for both integral fea-
tures and derived features.
We typically use angularity tolerances as a supplement to position tolerances
to ensure better component geometry than what these tolerances do on their
own.
Note that for the angularity tolerances, the coordinate system of the tolerance
zone is rotated relative to the coordinate system of the datum system. This
means that there is sometimes more than one datum participating in locking
a given degree of freedom for the tolerance zone.

Angularity example 1: Angularity of a plane surface

Figure 4–32 shows an example of an angularity tolerance for a real
plane surface (an integral feature).
The angularity tolerance applies to the entire plane surface.

Figure 4–32:  Angularity tolerance for a plane surface

relative to a datum system

This example shows an important detail where angularity tolerances

differ from parallelism tolerances and perpendicularity tolerances.
Because the angle to the primary datum can rotate around the normal
to the datum, it is necessary to reference a secondary datum to ensure
that the angle gets the intended orientation.

114 Find your way in GPS

Without the secondary datum, the 15° angle could be orientated from
side to side, or at an angle, instead of being orientated from end to end
of the toleranced feature.

Figure 4–33:  Explanation drawing for figure 4–32

If you compare figure 4–33 with figure 4–21, you can see that the an-
gularity tolerance and the parallelism tolerance with a plane as the da-
tum work the same way and have the same single unlocked degree of
freedom, as long as a datum for the angularity tolerance that locks the
orientation of the angle is indicated.

Find your way in GPS 115

 Angularity example 2: Angularity of the median line of a hole
Figure 4–34 shows an example of an angularity tolerance for the medi-
an line of a hole (a derived non-ideal feature). The angularity tolerance
applies to the entire feature.

Figure 4–34:  Angularity tolerance for the median line of a

hole relative to a datum system

This example, like the previous example, shows the necessity of a sec-
ondary datum to ensure that the angle gets the intended orientation.
Without the secondary datum, the 75° angle could rotate around the
normal of the primary datum N.

Location tolerance:
limits
A tolerance that
ation
the form, orient
ition)
and location (pos
of the feature.

116 Find your way in GPS

 Figure 4–35:  Explanation drawing for figure 4–34

If you compare figure 4–35 with figure 4–31, you can see that the angu-
larity tolerance and the perpendicularity tolerance with a plane as the
datum work the same way and have the same two unlocked degrees of
freedom, as long as a secondary datum for the angularity tolerance that
locks the orientation of the angle is indicated.

Find your way in GPS 117

 Angularity example 3: Angularity of a plane surface
Figure 4–36 shows an example of an angularity tolerance for a real
plane surface (an integral feature). The angularity tolerance applies to
the entire plane surface.

Figure 4–36:  Angularity tolerance for plane surface

relative to a datum

This example shows a situation where the use of a secondary datum to

ensure that the angle gets the intended orientation is not necessary. Be-
cause the component, apart from the toleranced feature, has cylindrical
symmetry around datum E, it does not matter that the tolerance zone
can rotate around datum E.

118 Find your way in GPS

 Figure 4–37:  Explanation drawing for figure 4–36

In the specific example, you could imagine that the angled surface
drives a valve lifter or a pump piston. The angularity tolerance ensures
that you achieve the correct stroke. A deviation in the axial position of
the angled surface would not influence the stroke, but would require an
adjustment of the valve lifter or the piston rod.

Find your way in GPS 119

 Position

As described above, you usually use position to create fixed tolerances. But
you can also use position to create mobile related tolerances, if the referenced
datums do not lock all non-redundant degrees of freedom for the tolerance
zone.

Position example 4: Position of a median line of a hole II

Figure 4–38 shows an example of a position tolerance for the medi-
an line of a hole (a derived non-ideal feature). The tolerance does not
control the geometry of the hole, only the location, orientation and
straightness.

Figure 4–38:  Position tolerance for the median line of a

hole relative to a datum, cylindrical tolerance zone

120 Find your way in GPS

 Figure 4–39:  Explanation drawing for figure 4–38

If we compare the position tolerance in figure 4–38 with the position

tolerance in figure 4–4, we will see that the only difference between the
two position tolerances is that secondary datum A is not referenced in
figure 4–38.
By comparing figure 4–39 and figure 4–5 we can see that this difference
means that the tolerance zone gets two unlocked degrees of freedom,
because the distance to datum A and the orientation relative to datum
A is no longer locked.
This comparison illustrates how we can change a position tolerance
from a fixed tolerance to a mobile tolerance by changing which datums
we reference.

Find your way in GPS 121

 If instead we compare the position tolerance in figure 4–38 with the
parallelism tolerance in figure 4–24, we can see the difference in mean-
ing between the two symbols.
Position locks both the distance and the orientation relative to the ref-
erenced datums.
Parallelism (as well as perpendicularity and angularity) only locks the
orientation. So where the distance between the tolerance zone and da-
tum B is locked in figure 4–39, this degree of freedom is unlocked in
figure 4–5.
This comparison illustrates how the position symbol signals that the refer-
enced datums lock as many degrees of freedom as possible, whereas the ori-
entation symbols, parallelism, perpendicularity and angularity, only let the
referenced datums lock rotational degrees of freedom.
This means that we have a set of tools that allows us to express very precisely
how a tolerance zone is locked and how it is allowed to move.
By combining these tools, you can create a set of requirements for a feature
that consists of tolerance zones with different shapes, width and unlocked
degrees of freedom and thus expresses very precisely and nuanced functional
requirements for the feature.

Surface profile

As described above, you can use the surface profile symbol to create fixed tol-
erances. But just like the position symbol, you can also use the surface profile
symbol to create mobile related tolerances, if the datums you reference do not
lock all the non-redundant degrees of freedom for the tolerance zone.
Like position, surface profile tolerances lock both distance and orientation
relative to the referenced datums.

Surface profile example 2: Mobile location tolerance for a real

surface feature
The tolerance in figure 4–40 is an example of a related mobile tolerance,
where the referenced datums do not lock all the non-redundant degrees
of freedom for the tolerance zone.

122 Find your way in GPS

The component is the same as in figure 4–10. The toleranced feature is
the periphery of the cam. As described above, we have to use the “all
around” symbol to indicate that the tolerance applies to the entire pe-
riphery.
We also have to use the CZ (Common Zone) modifier to indicate that
the tolerance zones for the four surfaces that make up the periphery are
locked together and have to move as one. In the previous surface profile
example, we did not have to use CZ because the tolerances were fixed,
so there was no question whether they moved individually or as a set.
It would not have changed the meaning, had we used the CZ modifier
in figure 4–10.

Figure 4–40:  Surface profile tolerance for a profile that

consists of 4 features with reference to a single datum

Find your way in GPS 123

 Figure 4–41:  Explanation drawing for figure 4–40

If we compare the tolerance in figure 4–40 with the tolerance in figure

4–10, we can see that the only differences are that the secondary datum
B is not referenced in figure 4–40 and that we have added the CZ modi-
fier.
By comparing figure 4–41 to figure 4–11 we can see that this difference
means that the tolerance zone gets one unlocked degree of freedom,
because the orientation relative to datum B (the keyway) is no longer
locked.
You could imagine using a combination of these two tolerances, where
you would use a tighter tolerance (smaller tolerance value) in figure
4–40 to control the geometry of the cam periphery and its alignment
relative to the hole in combination with a looser tolerance (higher tol-
erance value) in figure 4–10 to control the angular position of the cam
periphery relative to the keyway.
This comparison illustrates how we can change a surface profile toler-
ance from a fixed tolerance to a mobile tolerance by changing which
datums we reference.

124 Find your way in GPS

 The comparison also shows how we can use a combination of toler-
ances with the same symbol, but with different numbers of datums and
different tolerance values to control different functional properties for
a feature.
When you use this kind of combination tolerances, you have to make sure
that the datums used in the tolerance with the fewest datums are also used
in the same sequence in the tolerance that uses the most datums. This means
that you have to remove datums from the rear of the datum system, when you
want to unlock degrees of freedom for the tolerance zone.

Line profile

As described above, you can use the line profile symbol to create fixed toler-
ances. But like the surface profile symbol and the position symbol, you can
also use the line profile symbol to create mobile related tolerances, if the ref-
erenced datums do not lock all the non-redundant degrees of freedom for the
tolerance zone.
Line profile tolerances lock both distance and orientation relative to the ref-
erenced datums, just like surface profile and position.

Find your way in GPS 125

 Line profile example 2: Mobile location tolerance for a real sur-
face feature
The tolerance in figure 4–42 is an example of a related mobile tolerance,
where the referenced datums do not lock all the non-redundant degrees
of freedom for the tolerance zone.
The component is the same as in figure 4–10 and 4-12. The toleranced
feature is the periphery of the cam. As described above, we have to use
the “all around” symbol to indicate that the tolerance applies to the en-
tire periphery.
We also have to use the CZ (Common Zone) modifier to indicate that
the tolerance zones for the four surfaces that make up the periphery are
locked together and have to move as one. We did not have to use CZ in
the previous line profile example, because the tolerances were fixed, so
there was no question whether they moved individually or as a set. It
would not have changed the meaning, had we used the CZ modifier in
figure 4–12.
In figure 4–12 we saw that when all non-redundant degrees of freedom
are locked, there is no difference between a surface profile tolerance
and a line profile tolerance.
But when we have unlocked degrees of freedom, each line tolerance
zone can move relative to the other tolerance zones, so the line profile
tolerance becomes less restrictive than the corresponding surface pro-
file tolerance.

Figure 4–42:  Line profile tolerance for a profile that con-

sists of 4 features with reference to a single datum
126 Find your way in GPS
 ly

Figure 4–43:  Explanation drawing for figure 4–42

Figure 4–42 and figure 4–40 are identical, except that the surface pro-
file symbol is used in figure 4–40 and the line profile symbol is used in
figure 4–42.
By comparing figure 4–43 and figure 4–41 we can see the difference
between the two tolerances, when everything else is alike and there is
one or more unlocked degrees of freedom.
The difference turns out to be that where the surface profile tolerance
zone only can move as a unit, each line profile tolerance zone can move
independently of the others in the set. This creates an extra degree of
freedom. You can imagine the tolerance zones rotated relative to each
other, as when you fan out a deck of cards. This means that the cam
can be twisted as in figure 4–44 and still pass the line profile tolerance,
whereas surface profile does not allow this deviation.

Find your way in GPS 127

 Figure 4–44:  Geometric deviation for a cam that is al-
lowed by the line profile tolerance in figure 4–42, but not
by the surface profile tolerance in figure 4–40

You can imagine using a combination of the two tolerances, where you
would use a tighter tolerance (smaller tolerance value) in figure 4–42 to
control the geometry of the cam periphery and its alignment relative to
the hole in each cross-section, in combination with a looser tolerance
(larger tolerance value) in figure 4–40 to control how much the cam
profile is allowed to twist.
When you use such combination tolerances with a surface profile toler-
ance and a line profile tolerance, the tolerances shall reference the same
datums in the same sequence for the combination to work as intended.

4.3 Unrelated tolerance zones

Unrelated tolerance zones are tolerance zones that do not reference datums.
You can create unrelated tolerance zones with the dedicated form symbols
flatness, cylindricity, straightness and roundness. By definition you cannot
use datums with these symbols.
You can also use the surface profile and line profile symbols without reference
to datums.

128 Find your way in GPS

Flatness

Flatness is an unrelated form tolerance. “Unrelated” means that flatness toler-
ances do not use datums, so the tolerance zone can move freely and adjust to
the toleranced feature (the tolerance is not related to datums).
Flatness tolerances create surface requirements. This means that the entire
toleranced feature has to fit inside the tolerance zone simultaneously.
Flatness tolerances can be used for derived features, such as a median surface,
but are predominantly used for integral features (real surfaces).

Flatness example 1: Flatness of a real flat surface

Figure 4–45 shows an example of a flatness tolerance for a real flat sur-
face (an integral feature). The flatness tolerance applies to the entire
plane surface.

Figure 4–45:  Flatness tolerance for a flat surface

A flatness tolerance is limited by two perfect, parallel mathematical

planes and has three unlocked degrees of freedom. These three degrees
of freedom are the translatory motion up and down along the normal
to the plane and the rotations around the two axes in the plane. A flat-
ness tolerance zone can adjust to fit the toleranced feature by moving in
these three directions, see figure 4–46.

Find your way in GPS 129

 Orientation tolera
nce:
A tolerance that lim
its the
form and orientat
ion of the
feature, but not
its loca-
tion.

Figure 4–46:  Explanation drawing for figure 4–45

Flatness tolerances are typically used to ensure good contact between

the toleranced feature and its counterpart, often as a supplement to an
orientation tolerance like those in figures 4-21 and 4-26, or a location
tolerance like the one in figure 4–2.

130 Find your way in GPS

If you compare figure 4–46 to figure 4–22 and 4-27, you can see that the
difference between an orientation tolerance such as perpendicularity or
parallelism for a plane surface with a plane as the datum and a flatness
tolerance is that where the flatness tolerance zone has three unlocked
degrees of freedom, the orientation tolerance zones only have one or
two unlocked degrees of freedom. This means that with the same toler-
ance value, the orientation tolerances are more restrictive. Orientation
tolerances control form, but the flatness tolerance does not control the
orientation (perpendicularity or parallelism).

Flatness example 2: Flatness of a median surface

Figure 4–47 shows an example of a flatness tolerance for a median sur-
face (a non-ideal derived feature). The flatness tolerance applies to the
entire feature.

Figure 4–47:  Flatness tolerance for a median surface

In GPS we distinguish between ideal derived features that do not have

form deviations and non-ideal derived features that have form devia-
tions. For good reason it does not make sense to have a form tolerance
for an ideal feature, but sometimes it is necessary to use an ideal feature
to define a non-ideal feature. The median surface in this example is de-
fined as a set of points, each of which is the midpoint of a line between
two points, one on each surface. The problem in this definition is how
to decide which two points belong together. This is solved by requir-
ing that the line shall be perpendicular to the median plane, see figure
4–48.

Find your way in GPS 131

 Figure 4–48:  Definition of a median surface

132 Find your way in GPS

 Figure 4–49:  Explanation drawing for figure 4–47

It is very rare that a flatness tolerance is used for a median surface. The ex-
ample is only included for completeness.

Cylindricity

Cylindricity is an unrelated form tolerance. This means that cylindricity tol-
erances do not use datums, so the tolerance zone can move freely and adjust
to the toleranced feature.
Cylindricity tolerances create surface requirements for cylindrical features.
Cylindricity tolerances are used for integral features (real surfaces).

Find your way in GPS 133

 Cylindricity example: Cylindricity of a cylinder
Figure 4–50 shows an example of a cylindricity tolerance for a cylin-
drical integral feature. The cylindricity tolerance applies to the entire
feature.

Figure 4–50:  Cylindricity tolerance for a cylindrical

surface

A cylindricity tolerance is the space between two coaxial cylinders and

it has four unlocked motion degrees of freedom. Additionally, the toler-
ance zone can adjust to fit the toleranced feature by changing the me-
dian diameter of the two cylinders. So all in all a cylindricity tolerance
has five unlocked degrees of freedom.

Figure 4–51:  Explanation drawing for figure 4–50

134 Find your way in GPS
 A cylindricity tolerance ensures the roundness of each cross-section of
the cylinder, the straightness of the generatrices, the straightness of the
cylinder axis, and the diameter variation of the cylinder. Cylindricity
requirements are typically used where components have to move rela-
tive to each other, either axially, or in a rotation, or both.
Cylindricity tolerances ensure that the fit between two components re-
main consistent when they move relative to one another. Examples of
such components are engine cylinders and pistons, hydraulic cylinders,
air cylinders and fuel injection systems.

Surface profile

As described above, you can use the surface profile symbol to create fixed and
mobile related tolerances. But you can also use the surface profile symbol to
create unrelated form tolerances.
You can think of cylindricity and flatness as special cases of surface profile.
In order to use a surface profile tolerance for a non-plane surface, you have
to define the nominal geometry of the surface. You can for example do this
by using TEDs. Other possibilities are to define the surface using spline func-
tions or a table of coordinates and an interpolation function between these
coordinates.

Surface profile example 3: Unrelated form tolerance for a real

surface feature
The tolerance in figure 4–52 is an example of an unrelated form toler-
ance, where the tolerance zone can move freely to fit to the toleranced
feature. In contrast to the cylindricity tolerance zone that can change
the median diameter, the shape and size of the surface profile tolerance
zone are locked by the nominal geometry.
As in figure 4–40 we have to use the CZ (Common Zone) modifier to
indicate that the tolerance zones for the four surfaces that make up the
periphery are locked together and have to move as one.

Find your way in GPS 135

 Figure 4–52:  Unrelated surface profile tolerance for a
profile that consists of 4 features

Figure 4–53:  Explanation drawing for figure 4–52

136 Find your way in GPS

 The tolerance in figure 4–52 could be used to supplement the tolerance
in figure 4–10 and/or the one in figure 4–40. Because it has more un-
locked degrees of freedom, it has to have a smaller tolerance value than
the other two tolerances in order to create an additional requirement.
If you combined the three toler-
ances, the unrelated tolerance in lerance:
figure 4–52 would control the Mobile (related) to
shape and size of the feature. some, but
A tolerance where
The mobile related tolerance in ndant de-
not all, non-redu
figure 4–40 with a large toler- are locked
grees of freedom
ance value would control the datums
location and orientation of the with references to
cam profile relative to the hole in
the cam. Finally, the fixed toler-
ance in figure 4–10 with an even
larger tolerance would control
the orientation of the cam pro-
file relative to the keyway.

Straightness

Straightness is an unrelated form tolerance. This means that straightness tol-
erances do not use datums, so the tolerance zone can move freely and adjust
to the toleranced feature.
You can use straightness tolerances for
both integral features as in examples 1, 2
Unrelated tolerance: and 4 and for derived features, as in exam-
ple 3. A straightness tolerance always cre-
A tolerance that does ates a line requirement. If the toleranced
not reference datums feature is a surface, you can either apply
the requirement to all lines in the surface
in a certain direction, as in example 1, or
only to a specific line in the surface, as in
example 2.

Find your way in GPS 137

 Straightness example 1: Straightness of a flat surface
Figure 4–54 shows an example of a straightness tolerance for a surface
on a block. The straightness tolerance applies to all lines in the surface
parallel to the drawing plane for the picture that the leader line points
to. This illustrates two general rules:
1 When you create a line tolerance for a surface, it applies not just
for one line, but for all lines in the surface, unless you identify a
specific line (see example 2).
2 If the toleranced feature is a flat surface, the straightness tolerance
always applies in the direction of the paper plane.

Figure 4–54:  Straightness tolerance for a flat surface

138 Find your way in GPS

 Figure 4–55:  Explanation drawing for figure 4–54

The straightness tolerance thus defines a set of tolerance zone. Each tol-
erance zone can adjust to fit its own line in the surface independently of
the other tolerance zones by moving up and down and rotating around
the normal of the projection plane. The individual straightness toler-
ance zone thus has two unlocked degrees of freedom.
Even though this is the classic example of a straightness tolerance, it is
hard to imagine a component function where it is the straightness and
not the flatness of the surface that is relevant.

Find your way in GPS 139

 Straightness example 2: Straightness of one cross-section of a
flat surface
Figure 4–56 shows an example of a straightness tolerance for one iden-
tified cross-section on a surface on a block. The straightness tolerance
only applies to the identified cross-section, because the indication is
shown in a sectional view, as opposed to the tolerance in figure 4–54.

Figure 4–56:  Straightness tolerance for one cross-section

of a flat surface

The tolerance only defines a single tolerance zone because the straight-
ness tolerance only applies for one identified cross-section in the toler-
anced surface. The tolerance zone can freely adjust to the cross-section
in the surface by moving up and down and rotating around the normal
of the section plane.
This type of tolerance can, for example, be used when another com-
ponent contacts the surface exactly in this section and contact in the
entire section is necessary. In this case the tolerance can be used as
a supplement to a location and/or orientation tolerance for the entire
surface.

140 Find your way in GPS

Figure 4–57:  Explanation drawing for figure 4–56

Find your way in GPS 141

 Straightness example 3: Straightness of the median line of a
shaft
Figure 4–58 shows an example of a straightness tolerance for the me-
dian line of a shaft, which is a derived feature.

Figure 4–58:  Straightness tolerance for the median line of

a shaft

In this case there is an ¤ in front of the tolerance value, as opposed to

the previous two examples. This tells us that the tolerance zone is cy-
lindrical.
Another difference is that the leader line from the tolerance indicator
ends at a dimension line. It is a general rule that the tolerance applies
to the derived feature (the median line) when the leader line ends at a
dimension line. It is easy to overlook this rule, because if you forget it,
there is nothing in the drawing to remind you of it.

142 Find your way in GPS

 Figure 4–59:  Explanation drawing for figure 4–58

The tolerance zone has four unlocked degrees of freedom. These four
degrees of freedom are the two translatory movements perpendicular
to the axis and the rotations around these two cross axes. The cylindri-
cal straightness tolerance zone can adjust to fit the real median line by
moving in these four directions.
This type of tolerance can be used when you want to ensure that the
shaft is not banana shaped or corkscrew shaped. However, this tol-
erance does not ensure that the shaft is round or that it is not barrel
shaped or bell mouthed. If you also want to ensure this, you have to
either add a roundness tolerance and/or a straightness tolerance to the
generatrices of the shaft, or a cylindricity tolerance that controls every-
thing on its own.
As always you have to make sure that you make requirements for all
those characteristics that are important, so you assure that the compo-
nent will function, but not for the characteristics that are not impor-
tant, so the component does not become unnecessarily expensive to
produce.

Find your way in GPS 143

 Straightness example 4: Straightness of the generatrices of a
shaft
Figure 4–60 shows an example of a straightness tolerance for the gen-
eratrices of a shaft, which are integral features.

Figure 4–60:  Straightness tolerance for the generatrices of

a shaft

This example, just like the first straightness example, applies to every
line in the surface, and each tolerance zone can adjust to its generatrix
independently of the other tolerance zones. In contrast to the first ex-
ample, here it is self-evident in which direction the requirement applies.

Figure 4–61:  Explanation drawing for figure 4–60

144 Find your way in GPS

 This type of tolerance indirectly ensures the straightness of the median
line of the shaft, because it is impossible for all the generatrices to be
straight, if the median line is not straight as well. On the other hand,
this tolerance does not ensure the shaft being round or not being thick-
er at one end than at the other looking like a Christmas tree. If you
also want to control these characteristics, you have to add a roundness
tolerance and a parallelism tolerance, or a cylindricity tolerance that
controls everything on its own.

Roundness

Roundness is an unrelated form tolerance. This means that roundness toler-
ances do not refer to datums, so the tolerance can freely move and adjust to
fit the toleranced feature.
A roundness tolerance always creates line requirements and, apart from
rare exceptions, roundness tolerances are only used for cylinders, cones and
spheres. A roundness requirement can either apply to all cross-sections per-
pendicular to the axis (or the centre point of a sphere), or to one specific
cross-section of the feature.
The direction of the tolerance zone for roundness is always perpendicular to
the axis. Roundness is the only requirement that is not by default perpendicu-
lar to the surface. The result is the same for a cylinder, see example 1 below,
but different for a cone, see example 2.

Roundness example 1: Roundness of a cylinder

Figure 4–62 shows an example of a roundness tolerance for the surface
of a cylinder. The roundness tolerance applies to all cross-sections of
the cylinder perpendicular to the axis. The tolerance value is the ra-
dius difference between the two concentric circles limiting the toler-
ance zone. Therefore if a component is oval, the difference between the
smallest and largest two-point diameter can be up to twice the round-
ness tolerance.

Find your way in GPS 145

 Default:
unless
Rules that apply
. For
otherwise specified
untries
example, in some co
t speed
50 km/h is the defaul
limit in cities.

Figure 4–62:  Roundness tolerance for a cylinder

Figure 4–63:  Explanation drawing for figure 4–62

The roundness tolerance ensures that the cylinder is round in each

cross-section. It does not ensure that the diameter is constant, so the
cylinder can be barrel shaped, trumpet shaped or look like a Christmas
146 Find your way in GPS
tree. If this also has to be ensured, you have to use a cylindricity toler-
ance instead. Neither does it ensure that the median line is not banana
shaped or looks like a corkscrew. If this also has to be ensured, you have
to add a straightness tolerance to the generatrices or to the median line.

Roundness example 2: Roundness of a cone

Figure 4–64 shows an example of a roundness tolerance for the surface
of a cone. The roundness tolerance applies to all cross-sections of the
cone perpendicular to the axis. As mentioned above, this is the only
example where the width of the tolerance zone is not by default perpen-
dicular to the surface.

Figure 4–64:  Roundness tolerance for a cone

Figure 4–65:  Explanation drawing for figure 4–64

Find your way in GPS 147
 The roundness tolerance ensures that the cone is round in each cross-
section. It does not ensure that the generatrices are straight, so the cone
can look like a hemisphere or the mouth of a trumpet. If this also has to
be ensured, you have to add a straightness tolerance to the generatrices.
Neither does it ensure that the median line is not banana shaped or
looks like a corkscrew. If this also has to be ensured, you have to add
a straightness tolerance to the median line itself or to the generatri-
ces, because they cannot all be straight without the median line being
straight.
Finally, the roundness tolerance does not ensure that the cone angle is
correct.

Roundness example 3: Roundness of a cross-section of a hole

Figure 4–66 shows an example of a roundness tolerance for an identi-
fied cross-section of a cylindrical hole. The roundness tolerance only
applies to the identified cross-section.

Figure 4–66:  Roundness tolerance for an identified

cross-section of a hole

A roundness tolerance for a single cross-section is only used if the

cross-section is special, for example if another component has to fit into
the hole in exactly this depth and maybe seal with an O-ring.

148 Find your way in GPS

 -

Figure 4–67:  Explanation drawing for figure 4–66

Line profile

As described above, you can use the line profile symbol to create both fixed
and mobile related tolerances. But you can also use the line profile symbol to
create unrelated form tolerances.
You can think of roundness and straightness as special cases of line profile.
In order to be able to use line profile tolerances for a nominally non-flat sur-
face or a nominally non-straight line, it is necessary to define the nominal
geometry of the surface or the line. You can do this with TEDs.

Find your way in GPS 149

 Line profile example 3: Unrelated form tolerance for a real sur-
face feature
The tolerance in figure 4–68 is an example of an unrelated form toler-
ance, where the tolerance zone can move freely and adjust to fit the
toleranced feature. In contrast to the roundness tolerance zones that
can change median diameter, the shape and size of the surface profile
tolerance zone are locked by the nominal geometry.
As in figure 4–52 we have to use the CZ (Common Zone) modifier to
indicate that the tolerance zones for the four surfaces that make up the
periphery are locked together and have to move as one.

Figure 4–68:  Unrelated line profile tolerance for each

cross-section parallel to the paper plane for a profile that
consists of 4 features

Figure 4–68 and figure 4–52 are identical, with the exception that the
line profile is used in figure 4–68, and the surface profile symbol is used
in figure 4–52.
By comparing figure 4–69 and figure 4–53, we can see the difference
between the two tolerances, when everything else is the same and there
is one or more unlocked degrees of freedom.
The difference turns out to be that where the surface profile tolerance
zone only can move as a unit, each line profile tolerance zone can move
independently of the others in the set. This creates an extra set of degree
of freedom. You can imagine the tolerance zones as not only rotated

150 Find your way in GPS

relative to each other, as in figure 4–44, but also displaced relative to
each other in the two directions in the paper plane.
You can imagine using a combination of the two tolerances, where you
would use a tighter tolerance (smaller tolerance value) in figure 4–69
to control the geometry of the cam periphery in each cross-section in
combination with a looser tolerance (larger tolerance value) in figure
4–52 to control how much the cam profile is allowed to be offset and
twisted.
In principle you could combine all the tolerances in figure 4–10, figure
4–12, figure 4–40, figure 4–42, figure 4–52 and figure 4–68 and express
some very specific requirements for the form, orientation and location
of the cam profile. However, in reality you would only use two or three
of the tolerances, because the differences between the tolerance values
would otherwise be too small to gain any advantage by separating the
requirements.

Figure 4–69:  Explanation drawing for figure 4–68

Find your way in GPS 151

 4.4 Run-out tolerances
Run-out tolerances are a special type of tolerances that combine form require-
ments for the integral feature (the real surface) with location and orientation
requirements for the derived feature.
We talk about radial run-out when the run-out requirement is indicated per-
pendicular to the datum axis and axial run-out when it is indicated in the
direction of the datum axis.
Radial run-out always has the median diameter of the tolerance zone as the
only unlocked degree of freedom and axial run-out always has the position
along the datum axis as the only unlocked degree of freedom.
The total run-out symbol creates surface requirements. You can use total run-
out for cylindrical features that are coaxial with the datum axis and flat sur-
faces that are perpendicular to the datum axis.
You can think of total run-out for a cylindrical feature as cylindricity relative
to a datum axis and total run-out for a flat surface as perpendicularity.
The circular run-out symbol creates a set of line requirements for each line in
the toleranced feature. You can use circular run-out for cylindrical features,
conical features, trumpet shaped features and barrel shaped features that are
coaxial with the datum axis as well as flat surfaces that are perpendicular to
the datum axis.
You can think of circular run-out for a cylindrical feature as roundness rela-
tive to a datum axis.

Total run-out

You can use the total run-out symbol to create related surface requirements.
You always have to indicate total run-out with an axis as the datum. The
toleranced feature can either be cylindrical and coaxial with the datum axis
or flat and perpendicular to the datum axis.

152 Find your way in GPS

Total run-out example 1: Related surface requirement for a
real, cylindrical surface feature
The tolerance in figure 4–70 is an example of a radial total run-out tol-
erance for a cylindrical feature. The tolerance zone is fixed coaxially
with the datum axis and the tolerance zone can only adjust to fit the
toleranced feature by changing the median diameter. The tolerance can
be interpreted as cylindricity relative to a fixed axis.
The tolerance refers to a common datum. We will cover datums and the
meaning of a common datum in chapter 5.

Figure 4–70:  Total run-out tolerance for a cylindrical

feature with reference to a common datum

Find your way in GPS 153

 Total run-out:
al de-
d effect of several geometric
Total run-out is the combine
for fea tu res wit h cy lindr ica l symmetry considered for
viations lue
e en tire fea tu re sim ult an eo usly. You can say that the va
th m.
“runs out” relative to the datu
expresses how far the feature de-
typically used to limit these
Total run-out tolerances are
in rot at ing co mp on en ts . There are two kinds of total
viations
run-out tolerances:
e fea-
e radius variation in the entir
Radial total run-out limits th mi-
as cylindricity relative to a no
ture and can be interpreted
ial da tu m line. Ra dia l to ta l run-out can therefore only
nally coax
.
be used for cylindrical features
hed
ial to ta l run -ou t co nt rol s wobbling (like a loosely attac
Ax a-
n for the entire feature simult
car wheel) in the axial directio tum
rpendicularity relative to a da
neously and is the same as pe
ial to ta l run -ou t ca n th ere fore only be used for plane sur-
line. Ax
ed by perpendicularity.
faces and can always be replac

154 Find your way in GPS

 Figure 4–71:  Explanation drawing for figure 4–70

You typically use radial total run-out tolerances to control a rotor rela-
tive to its bearing surfaces. Especially if the rotor has to fit into a hous-
ing where there are tight tolerances between the rotor and the housing.
An example of this is an electric motor, where the efficiency among
other things depends on how small you can make the gap between the
rotor and the stator.

Find your way in GPS 155

 Total run-out example 2: Related surface requirement for a
real, flat surface feature
The tolerance in figure 4–72 is an example of an axial total run-out tol-
erance for a flat feature. The tolerance zone is oriented perpendicular
to the datum axis. The tolerance zone can adjust to fit the toleranced
feature by moving linearly along the datum axis. A perpendicularity
tolerance would have the exact same meaning.

ing

Figure 4–72:  Total run-out tolerance for a flat feature with

reference to a datum

Figure 4–73:  Explanation drawing for figure 4–72

156 Find your way in GPS

Circular run-out

You use the circular run-out symbol to create related line requirements. You
always have to indicate circular run-out with an axis as the datum. The toler-
anced feature can either have cylindrical symmetry (e.g. cylindrical, conical,
trumpet shaped or barrel shaped) and coaxial with the datum axis or flat and
perpendicular to the datum axis.

Radial circular run-out example: Related line tolerances for a

set of lines in a real, cylindrical surface feature
The tolerance in figure 4–74 is an example of a radial circular run-out
tolerance for a cylindrical feature. The tolerance zones are fixed coaxi-
ally with the datum axis and each tolerance zone can adjust to fit in-
dividually to its toleranced feature by changing its median diameter
independently of the other tolerance zones. The tolerance can be inter-
preted as roundness relative to a fixed axis.
The tolerance refers to a common datum. We will cover datums and the
meaning of a common datum in chapter 5.

Figure 4–74:  Circular run-out tolerance for a cylindrical

feature with reference to a common datum

Find your way in GPS 157

 -

Figure 4–75:  Explanation drawing for figure 4–74

The difference between radial circular run-out tolerances and radial

total run-out tolerances is that radial total run-out does not allow dia-
meter variation along the axis, so it can only be used for cylindrical
features. Radial total run-out ensures that the feature is cylindrical,
whereas a conical, trumpet shaped or barrel shaped feature can pass a
radial circular run-out (not total) tolerance.

158 Find your way in GPS

Axial circular run-out example: Related line tolerances for a set
of lines in a real, flat surface feature
The tolerance in figure 4–76 is an example of an axial circular run-out
tolerance for a flat feature. The tolerance zones are oriented perpen-
dicular to the datum axis. Each tolerance zone can adjust to fit the toler-
anced feature by moving linearly along the datum axis.

Figure 4–76:  Circular run-out tolerance for a flat feature

with reference to a datum

Find your way in GPS 159

 -

Figure 4–77:  Explanation drawing for figure 4–76

The difference between axial circular run-out tolerances and axial total
run-out tolerances is that axial total run-out does not allow the toler-
anced feature to be dish shaped, i.e. conical, either outwards or inwards.
Axial circular run-out and not axial total run-out is the correct require-
ment in many cases where a cylindrical component shall rest on its end
surface. In this case the requirement shall be supplemented with the
modifier NC, which means “non-convex”. This means that the end sur-
face is allowed to be hollow (or concave), but not to bulge outwards.
This ensures that the component can rest on the rim of the surface in a
stable manner.

160 Find your way in GPS

 Circular run-out:
ct of several geometrical de-
Circular run-out is the combined effe
l symmetry considered in one
viations for features with cylindrica
to total run-out. You can say
cross-section at a time, as opposed
feature “runs out” in each
that the value expresses how far the
. Circular run-out tolerances
cross-section relative to the datum
iations in rotating compo-
are typically used to limit these dev
r run-out tolerances:
nents. There are two kinds of circula
ius variation in one cross-
Radial circular run-out limits the rad
ed as roundness relative to
section at a time and can be interpret
l circular run-out allows ra-
a nominally concentric datum. Radia
and can therefore be used for
dius deviation in the axial direction
radial total run-out.
e.g. conical features, as opposed to
g in the axial direction of one
Axial circular run-out controls wobblin
radial distance) at a time, re-
concentric cross-section (i.e. in one
run-out allows dish shape, as
lative to a datum line. Axial circular
pendicularity), but ensures
opposed to axial total run-out (or per
dicular to the datum. It can
that the dish is centred and perpen
es.
therefore be used for non-plane surfac

Find your way in GPS 161

 4.5 Summary
In this chapter we have covered the 14 tolerance symbols and looked at some
typical uses.
We have seen that the GPS tolerances can be classified in different ways. There
are related tolerances that reference datums, and the unrelated, or form toler-
ances, that do not.
Among the related tolerances we distinguished between the fixed tolerances
that have all their non-redundant degrees of freedom locked and the mobile
tolerances that have one or more unlocked degrees of freedom.
We have seen that some symbols always create surface tolerances, others al-
ways create line tolerances and yet others can create either, depending on the
context.
We have seen that you can indicate if a tolerance applies to the identified inte-
gral feature or its derived feature, depending on whether you place the leader
line away from the dimension line or aligned with the dimension line.
We have seen that the fixed tolerances that have all their non-redundant de-
grees of freedom locked are the only ones that can stand alone. We create
these tolerances with the position, symmetry, coaxiality, surface profile, and
line profile symbols.
We have discussed that the orientation symbols parallelism, perpendicular-
ity and angularity shall reference datums, but only lock rotational degrees of
freedom, regardless of whether the referenced datums would be capable of
also locking some of the translatory degrees of freedom.
We have seen that the dedicated form symbols can never be used with da-
tums.
Finally, we have seen that the surface profile and line profile symbols are very
flexible and can be used without datums to create form tolerances or with da-
tums to create fixed or mobile tolerances, depending on which and how many
datums are referenced in the tolerance indicator.

162 Find your way in GPS

5 Datums and datum systems

The fundamental problem with dimensional tolerances, as we saw in chapter

2, is that they are not tied together in a coordinate system, which means that
we cannot ensure that the height and width of a component are perpendicu-
lar to each other, even though it may look that way in the drawing, see figure
5–1.

Figure 5–1:  Dimensional tolerances are not tied together

in a coordinate system

In chapter 4 we covered the 14 tolerance symbols. We also looked at a number

of examples showing how individual tolerance zones can be used to create re-
quirements for the features in a component. In many of the examples we used
datums without discussing how datums, that are perfect features, are derived
from (fitted to) the real datum feature that has form and angular errors.
In this chapter we will cover how datums are derived from datum features
and how you can call upon several datums in the tolerance indicator to refer-
ence a datum system.
Figure 5–2 shows an example of how we can tolerance the component in fig-
ure 5–1 to ensure that the sides of the component are perpendicular to each
other and are in the correct distances by referencing datums in the tolerance
indicator and indicating the nominal distances.

Find your way in GPS 163

 Figure 5–2:  Geometrical tolerances can be tied together in
a coordinate system

In figure 5–2 the perpendicularity tolerance underlined in red (there are no

underlines in real drawings) ensures that the right side of the component is
perpendicular to the bottom surface indicated as datum A.
The position tolerance underlined in green ensures that the top surface of
the component is parallel to the bottom surface and is located at the correct
distance.
Finally, the position tolerance underlined in blue ensures that the left side of
the component is perpendicular to the bottom surface (because it first refer-
ences datum A) and is in the correct distance from the right side of the com-
ponent (because it references datum B and the nominal distance is indicated).
In this way the tolerance zones for all the surfaces in the component are
locked together in a coordinate system, where they are all either parallel or
perpendicular to each other and are locked in the correct distances, so we do
not risk the same problems as in figure 5–1.
The other problem we avoid is that it is not always clear from where and to
where a dimensional tolerance applies. We looked at some of these problems
in figures 2-18 and 2-19. With geometrical tolerances the toleranced feature
has to remain within the tolerance zone and there are clear rules for from
where and in what direction the tolerance applies. These rules are based on
datums and datum systems and this is what we are going to cover in this
chapter.

164 Find your way in GPS

5.1 The coordinate system of the component
The major advantage of geometrical tolerancing over dimensional toleran-
cing is that geometrical tolerances exist in the theoretically perfect nominal
world and that it is possible to lock all the tolerances together in a coordinate
system in this nominal world.
This requires a way to express how the component shall fit into this perfect
nominal world. This is where datums come into the picture.
We use datums to identify which features in the component we want to use
to define how the component shall be fitted into the coordinate system in the
nominal world.
The word datum comes from surveying, where it defines the reference from
which a measurement is taken. In GPS we talk about individual datums de-
rived from datum features and datum systems that are sets of datums, which
serve as the coordinate system for the tolerance zones and the tolerancing of
the component.
Every component has 6 degrees of freedom as shown in figure 5–3: Transla-
tion (linear motion) along the X, Y and Z axes and rotation around each of
the three axes.

Figure 5–3:  The 6 degrees of freedom

As a general rule, a datum system shall lock these 6 degrees of freedom. There
can be situations where one or more degree of freedom is redundant (without
meaning) for a given tolerance zone.

Find your way in GPS 165

 A datum system consists of a primary (first) datum, a secondary (second)
datum and, in some cases, a tertiary (third) datum. This prioritisation means
that the primary datum first locks as many degrees of freedom as it can. Next,
the secondary datum locks as many degrees of freedom not already locked by
the primary datum as it can. Finally, the tertiary datum locks as many degrees
of freedom as it can of all the degrees not already locked by the primary or
secondary datum.

5.2 Datums and datum features

We make a distinction between the datum feature, which is a real feature (a
surface on the component) with form and angular errors and the datum itself,
which is a geometrically perfect derived feature. Derived features are virtual
features that do not exist physically. Figure 5–4 shows the difference between
the datum feature and the datum. A datum is always either a plane, a straight
line or a point.

(derived feature with perfect geometry)

Figure 5–4:  The real datum feature and the derived datum

Figure 5–5:  A datum is a point, a straight line or a plane

The GPS system is primarily designed to ensure that components can be as-
sembled and that all the features of all the components are placed correctly
relative to each other after assembly. This is why the datum is derived from
the datum feature according to a set of rules simulating how the datum fea-
ture would be oriented and positioned, if assembled against a perfect feature

166 Find your way in GPS

with the same nominal geometry as the datum feature itself. For example, if
the datum feature is a nominally plane surface on the component, the derived
datum would be a mathematically ideal plane.

Planes as datums
A good way to imagine how a component can rest in a stable manner against
a plane surface is to imagine how a stool stands on a floor. A three-legged
stool never rocks. This is because three points always form a plane, so the
three-legged stool is stable.
A four-legged stool, on the other hand, will always rock, unless the floor is
perfectly flat and all four legs are the exact same length. If we have a four-
legged stool where the left rear leg and the right front leg are longer than the
other two legs, the stool will rock between the two stable positions resting on
the two long legs and one of the others. If our stool had more legs, it would
have even more stable positions.

Figure 5–6:  A three-legged stool is stable, a four-legged

stool rocks

We can think of a surface that is intended to be flat as a stool with an infinite

number of legs – one in each point on the surface. A real surface is never
perfectly flat. It has roughness and form error. This means that the legs of our
stool do not have the same length. This means that the surface will contact a
perfect plane in three points, just like the stool.
The contact with the datum plane is stable and the datum plane is well de-
fined, if the three highest points (the three longest legs) are far apart and the
triangle they form contains the centre point of the surface. Figure 5–7 shows
a 2-dimensional example with a line that rests on two points, but the principle
is the same for a 3-dimensional plane that requires three points to be stable.

Find your way in GPS 167

 (derived feature with perfect geometry)
Figure 5–7:  The datum plane for a nominally plane datum feature
is the plane upon which the feature rests on in a stable manner
with contact in the highest points, if this is possible

If the datum feature is a plane surface, the datum plane is associated with (fit-
ted to) the datum feature in such a way that the two touch each other in the
three highest points, if the datum feature is stable resting on the datum plane
in this manner.
If the datum feature is convex (bulging outward), you can imagine that it can
rest against the plane in several positions, each based on a different set of
three points, just like the four-legged stool can rock between the two stable
positions.
In this case the datum plane is associated thus minimising the largest distance.
ISO 5459:2011 specifies that the “holes” in the datum feature shall not be
taken into account when the distance is minimised. This fits logically with the
purpose of datums.

(derived feature with perfect geometry)

Figure 5–8:  If the datum plane for a nominally plane datum
feature is not stable, it is arranged such that the largest distance
between the datum and the datum feature is minimised, disre-
garding any “holes” in the datum feature

168 Find your way in GPS

Outside cylinders (shafts) as datums
The datum is the axis of the minimum circumscribed cylinder for the datum
feature if the datum feature is an outside cylinder (a shaft).
The minimum circumscribed cylinder is the smallest perfect cylinder that
will fit around the real cylinder. In other words, the smallest perfect hole the
shaft will fit into.

Figure 5–9:  When the datum feature is a shaft, the datum

is the axis of the minimum circumscribed cylinder

It is possible that the datum feature can rock in the minimum circumscribed
cylinder, for example if the datum feature (the real cylinder) is barrel shaped.
In this case the datum is oriented such that the largest distance between the
two is minimised.

Inside cylinders (holes) as datums

The datum is the axis of the maximum inscribed cylinder for the datum fea-
ture if the datum feature is an inside cylinder (a hole).
The maximum inscribed cylinder is the largest perfect cylinder (shaft) that
will fit inside the real cylinder.

Find your way in GPS 169

 Figure 5–10:  When the datum feature is a hole, the datum
is the axis of the maximum inscribed cylinder

There may exist more than one maximum inscribed cylinder for the datum
feature. In this case, the one that minimises the maximum distance between
the two is chosen.

5.3 Locking of degrees of freedom

A plane datum feature creates a plane datum that can lock the translation (the
linear movement) perpendicular to the plane and the two rotations, that are
not in the plane, see figure 5–11 compared to figure 5–3. Therefore you can
lock a total of three degrees of freedom with a datum plane.

Figure 5–11:  A datum plane can lock three degrees of

freedom

170 Find your way in GPS

A cylindrical datum feature (a shaft or a hole) creates a datum that is a straight
line. This kind of datum locks the two translations perpendicular to the line
and the two possible rotations of the line, see figure 5–12 compared to figure
5–3. Therefore you can lock a total of four degrees of freedom with a straight
line datum.

Figure 5–12:  A straight line datum can lock

four degrees of freedom

A spherical datum feature creates a datum that is a point. A point as a datum

locks all three translations and therefore three degrees of freedom, see figure
5–13 compared to figure 5–3.

Figure 5–13:  A datum point can lock three degrees of

freedom

Find your way in GPS 171

 In some cases more than one datum can be derived from a datum feature.
For example, if the datum feature is a cone, two datums can be derived: The
axis of the cone which is a straight line, and the apex of the cone, which is a
point on that line. This combination locks all degrees of freedom except the
rotation around this straight line (the cone axis). Therefore you can lock five
degrees of freedom with a conical datum feature.

Figure 5–14:  Two datums are derived from a cone, the axis
and the apex point

A wedge shaped datum feature creates two datums, a line and a plane. This
combination locks all degrees of freedom except the translation along the
width of the wedge. Therefore you can lock five degrees of freedom with a
wedge shaped datum feature.

5.4 Datum systems

You can lock a tolerance zone for one feature to another feature by identifying
the latter feature as the datum feature for the tolerance zone. You can also lock
the tolerance zone to a datum system by identifying several datum features.
The datum system works as a coordinate system for the tolerance zones.
The datums in the datum system are by definition parallel or perpendicular to
each other. This means that a secondary datum is constrained so that it is at
the perfect, nominal angle to the primary datum and a tertiary (third) datum
is constrained so that it is at the perfect, nominal angle to the primary datum
and the secondary datum. This means that the tolerance zones are located in
a perfect right angle coordinate system.

172 Find your way in GPS

 Datum:
at
Ideal derived feature th
fo r
is used as the basis
is
tolerancing. A datum
t
always a plane, a straigh
line or a point.

Figure 5–15:  The datum locks the tolerance zone and

therefore the toleranced feature to the datum feature
Find your way in GPS 173
 You find the functionally correct datum system by looking at how the com-
ponent is mounted in the overall product. There will always be a number of
features that control the position of the component in the product.
A datum system consisting of three mutually perpendicular planes is an ex-
ample of a datum system that can lock all six degrees of freedom.

Figure 5–16:  Alignment against the primary datum in the

datum system

The alignment against the primary datum happens as shown in figure 5–7
and 5-8, so the datum feature rests against the datum in the three highest
points, or so the largest distance is minimised.

Figure 5–17:  Alignment against the secondary datum in

the datum system

174 Find your way in GPS

The alignment against the secondary datum respects the alignment against
the primary datum. This means that the secondary datum by definition is
perpendicular to the primary datum and that it does not change the degrees
of freedom locked by the primary datum.
In the specific example this means that the rotation indicated by the arrow on
top of the component in figure 5–16 is already locked by the primary datum,
so the secondary datum must respect this locking and remain perpendicu-
lar to the primary datum. This means that the secondary datum feature will
touch the secondary datum in the two highest points, or the largest distance
is minimised, because the third degree of freedom is already locked.
It is a general rule that a plane datum feature will touch the datum plane in as
many points as there are degrees of freedom for it to lock, if it is stable.

Figure 5–18:  Alignment against the tertiary datum in the

datum system

The alignment against the tertiary datum respects the alignment against the
primary and secondary datums. This means that the tertiary datum by defini-
tion is perpendicular to both the primary and the secondary datums and that
it does not change the degrees of freedom locked by the primary and second-
ary datums.
In the specific example this means that the rotation indicated by the arrow on
one side of the component in figure 5–16 and the rotation indicated by the
arrow on the other side of the component in figure 5–17 are already locked
by the primary and secondary datums, so the tertiary datum must respect
this locking and remain perpendicular to the primary and secondary datums.

Find your way in GPS 175

 This means that the tertiary datum feature will touch the tertiary datum in
the single highest point, because the other two degrees of freedom are already
locked.
It is a general rule that when a plane datum locks one degree of freedom only,
the datum feature will only touch the datum plane in one point and there will
be no instability or rock to minimise.

5.5 Datum indicators

You use datum indicators to indicate the features you want to use as datums.
The datum indicator does not create any requirements by itself. You normally
use a single capital letter to identify each datum. Avoid using I, O and Q,
which can be read as numbers and therefore misunderstood. If you run out
of letters, you can use more than one letter in a datum indicator. It does not
matter which letters you choose, because the letters do not set the priority
amongst the datums in a datum system. This is governed by their sequence
when you use them in the tolerance indicator.

C K AA

Figure 5–19:  The datum indicator identifies a datum

feature with one or more capital letters – filling out the
datum triangle is optional

nt):
Real model (compone
ponent,
The physical com
angular
which has form and
on of a
errors. The verificati
easure-
component (the m
on the
ment) takes place
real component.

You identify the feature itself as the datum feature, if you attach the datum
triangle to a feature directly or to an extension line of the feature, but away
from the dimension line.

176 Find your way in GPS

 C

Figure 5–20:  Datum indicators identify the integral

feature as the datum feature unless the datum triangle is
aligned with a dimension line

You identify the derived feature as the datum feature, if you place the datum
triangle in extension of the dimension line. This means that you always have
to put the datum indicator in extension of the dimension line when you iden-
tify a feature of size as the datum feature.
It is especially in situations where a feature of size consists of two opposing
planes that are either parallel, so they form a width or a gap or have an angle
between them so they form a wedge, that the meaning is different depending
on whether the datum triangle is aligned with the dimension line or not.

B
A
C

Figure 5–21:  Datum indicators identify the derived

feature as the datum feature when the datum triangle is
aligned with a dimension line

5.6 Datum targets

Datum features consist by default of the entire identified feature. However, it
is possible to indicate that the datum feature consists of one or more parts of
the feature. In this case you have to indicate the area with a wide long dashed
dotted line and indicate the position of the area with non-toleranced dimen-
sions.

Find your way in GPS 177

 Figure 5–22:  If the datum feature is a partial feature, this
is indicated with a wide long dashed dotted line and non-
toleranced dimensions

Datum targets are another method you can use to indicate that the datum
feature is not made up of the entire feature. With datum targets you indicate a
number of areas on the feature that shall make up the datum feature.
Each target is identified by a circular datum target indicator with a horizontal
line through the middle. In the lower half you write the datum letter and se-
rial number for the target.
If the target is an area, you write the dimensions in the upper half, if space al-
lows. Otherwise you write the dimensions outside the indicator and connect
them with a reference line.
8×5
3 4 5 A1

4
A2

A A1, 2, 3
5,5
4

5,5
A3

Figure 5–23:  Indication of datum targets

178 Find your way in GPS
ISO 5459:2011 prescribes that a list of the datum targets that make up the da-
tum feature shall be indicated next to the datum indicator. This requirement
was not in the 1981 version of the standard. The idea is that it makes it easier
to ensure that you find all the targets involved.
It is a popular rule of thumb that you have to use three targets for a primary
datum feature, two for a secondary datum feature, and one for a tertiary da-
tum feature. As we saw earlier, these numbers coincide with the number of
degrees of freedom each datum locks, if they are planar. As we also saw, a
datum feature is stable, if it is resting on the same number of points as the
number of degrees of freedom it locks. This means that each datum feature
will be stable and uniquely resting on its datum targets, if you use these num-
bers of targets.
This rule comes from an idea that datum targets are used to define how mea-
surement fixtures shall be designed, for example for raw castings, so the com-
ponent will rest in a stable and unique (non-rocking) position on its datum
targets. This will give stable and repeatable measurement results, but it is not
necessarily functionally correct, if this is not how the component is fixed in
the product.
Because manufacturing of measurement fixtures is not the only use for datum
targets, it is absolutely possible to find situations where it makes functional
sense to use more targets. It can also make sense to use fewer targets, but in
this case you have to be quite certain that you know what you are expressing
with your drawing.

5.7 Common datum

You might also find yourself in the opposite situation where you want to use
more than one feature to define a datum. In this case you will use a common
datum.
The datum is associated with the indicated datum features simultaneously
and with equal weight, as if the datum features were one feature in a common
datum.

Find your way in GPS 179

 Figure 5–24:  Indication of a common datum

You indicate a common datum in the tolerance indicator by indicating the

datums that are part of the common datum separated by dashes. You can
indicate more than two datums in the common datum.
You typically use common datums in two situations. One is where the datum
features constitute an interrupted feature as in figure 5–25. The other is where
you use a group of similar features as in figure 5–25.
In the latter case you can show that you mean the group by writing ×N after
the datum indicator, where N is the number of features. You can either do it
with one datum indicator that indicates one of the features or with a datum
indicator that connects to a tolerance indicator that applies to all the features.
Of course you can also give each of the features its own datum indicator with
the same or different letters.
According to ISO 5459:2011 you shall repeat the letter separated by a dash
in the tolerance indicator when all the datum features have the same letter,
as in figure 5–25. This was not the case in the 1981 version of the standard.
Figure 5–25 follows the new version, which makes it clearer that it is a com-
mon datum.
If you indicate a number of parallel holes (or shafts) as a common datum, you
define two datums: The axis through the centre of the hole group and a plane
that contains this axis that locks the rotation around the axis.

180 Find your way in GPS

 Figure 5–25:  A group of features are indicated as a com-
mon datum (the TEDs necessary for a complete definition
of the tolerance zone are omitted)

Both in this case, where a common datum creates more than one datum and
in the case of a cone or a wedge, where one datum feature creates more than
one datum, you can find yourself in the situation where you need to express
that you are only interested in one of the created datums. In figure 5–25 the
datum D-D creates both a straight line (the centre line of the four holes) and
a plane that defines the orientation of the holes around the centre line. As the
line would lock the same degrees of freedom already locked by datum C (the
axis of the large hole), we are only interested in the plane.
However, the centre line for the four holes is automatically without conse-
quence, because datum D-D comes after datum C in the datum system, but if
you want to emphasise this, you can write (D-D)[PL] in the tolerance indica-
tor to emphasise that you only want to use the plane as a datum. There are
other situations where it makes a difference, if you indicate that you only want
to consider one datum from a datum feature or a common datum.

Find your way in GPS 181

 [PL] means that you only want to use a plane
[SL] means that you only want to use a straight line
[PT] means that you only want to use a point
In the case of a common datum, as here, you have to put parentheses around
the datum letters that identify the common datum. If it is only for an indi-
vidual datum, you do not have to write the parentheses.

5.8 Datums in the tolerance indicator

When you have named the datums with datum indicators, you are ready to
begin referencing them in tolerance indicators, either individually or as da-
tum systems. Datums are indicated in the third, fourth and fifth box in the
tolerance indicator.

Figure 5–26:  The tolerance indicator indicates an individual

datum that is used alone to constrain the tolerance zone

Figure 5–27:  The tolerance indicator indicates a prioritised

datum system that is used to constrain the tolerance zone

The sequence of the datums in a datum system matters, see figure 5–28 and
5-29. In figure 5–28 datum A is the primary datum and aligns the component
first. In figure 5–29 it is datum B that is the primary datum and aligns the
component first.
One of the consequences of this is that if you use datum B as an individual da-
tum for other tolerances in figure 5–28, then these tolerances will be aligned
differently from those tolerances aligned according to datum system |A|B|,
because secondary datum B by definition is perpendicular to primary datum
A, whereas the individual datum B only will be aligned by datum feature B.

182 Find your way in GPS

 Figure 5–28:  The tolerance indicator indicates that the
component is first aligned by datum A and then by datum B

Figure 5–29:  The tolerance indicator indicates that the

component is first aligned by datum B and then by datum A

You can and should lock all the tolerance zones you use to constrain the vari-
ous features in a component together by using the same datum system for all
the tolerance zones. This way you avoid wasting tolerance because the toler-
ance zones can move relative to each other. This enables you to use the larg-
est possible tolerances while still ensuring the function of the component.
You should always use such a global datum system, as it makes it cheaper to
manufacture the component.

Find your way in GPS 183

 If you feel tempted to use the secondary or tertiary datum from your global
datum system as an individual datum or change the sequence of your datums
in different tolerances, then resist the temptation and do not do it.
You shall always use datums from your global datum system in the same se-
quence to get the same alignment of all your tolerance zones. Just because the
datum features look perpendicular on the drawing, they will not be so on the
real component, see figures 5-28 and 5-29.
If you become unsure which features align the component in the product, you
have to go back and ensure that you have selected the correct global datum
system and then use it consequently for all the tolerance zones instead of
changing datum system halfway through the tolerancing.
Having said that, in some cases it will be functionally correct to use several
datum systems for a component. This can either be because a local group of
features in the component need to have tighter tolerances amongst them-
selves, or because your component can move in the product and uses differ-
ent features as interfaces in different positions. In these cases you will need
a local datum system for the tight tolerances or a number of equal datum
systems, one for each position.

5.9 Theoretically exact dimensions

Datums and datum systems provide us with the tools to indicate which fea-
tures of the component that form the reference (the starting point) for a toler-
ance. But we also need a tool that allows us to indicate the nominal distances
and sizes in the component. This tool is theoretically exact dimensions, or
TEDs.
20 25°

Figure 5–30:  Indication of theoretically exact dimensions

(TED)

A TED is indicated in the drawing in a rectangular box. It is indicated with

a dimension line just like all other dimensions. An angular TED is indicated
with the ° symbol and a length TED is simply indicated as a numerical value.
TEDs are nominal values that are used to lay out tolerance zones relative to
datums and each other. TEDs have no tolerances, and general tolerances (see
chapter 8) do not apply to TEDs.

184 Find your way in GPS

Length TEDs of 0 mm and angular TEDs of 0°, 90°, 180° and 270° do not have
to be indicated. In addition, an equal division of a circle does not have to be
indicated, so for example if five holes are shown equidistant on a circle, the
TED of 72° between them needs not to be indicated to apply.
So if you want to establish a position tolerance for a feature, you have to refer-
ence the datums from which the position has to be measured, but you also
have to make sure that there are TEDs (including those that do not need to be
indicated, as described above) connecting the datum features and the toler-
anced feature. You cannot use a toleranced dimension instead of a TED, be-
cause it does not exist in the nominal world and because it, in principle, does
not establish an exact dimension, only a pair of tolerance limits.
In chapter 4 we saw an example of what we call a TED pattern in figures 4-8
and 4-9 and we will look more into TED patterns in chapter 7. You create a
TED pattern by identifying a number of features, for example a number of
holes in a pattern, and by indicating their relative positions with TEDs (or
those TEDs not indicated) between them. You can then indicate a position
tolerance to all of the features relative to each other without indicating da-
tums as shown in figure 5–31.

Figure 5–31:  TED pattern

Find your way in GPS 185

 The TED pattern is a set of floating position tolerances that tolerance the in-
dicated features amongst themselves. In this way the features simultaneously
function as datums and toleranced features. Figure 5–32 shows the meaning
of figure 5–31.

The tolerance zones can adjust to fit the

toleranced features by moving, as a set,
vertically and horizontally and rotating around
all three axes in space. The tolerance zones are
not tied to the outer edges of the component

Figure 5–32:  Explanation drawing for figure 5–31

186 Find your way in GPS

We can reference datums for the TED pattern. It works just like a set of sepa-
rate tolerance zones, if we reference enough datums to completely lock the
pattern as in figure 4–8, but if we only reference datums that lock some, but
not all the degrees of freedom, for example by requiring that the holes have
to be perpendicular to the back face of the plate as in figure 5–33, then all the
tolerance zones still have to move together as a pattern, as in figure 5–32, but
they also have to respect the constraints from the referenced datums. Figure
5–34 shows the meaning of the tolerance in figure 5–33.

Figure 5–33:  TED pattern related to a datum

The TED pattern is a set of floating position tolerances that tolerance the in-
dicated features amongst themselves. In this way the features simultaneously
function as datums and toleranced features. Figure 5–32 shows the meaning
of figure 5–31.

Find your way in GPS 187

 The tolerance zones can adjust to fit the
toleranced features by moving, as a set,
vertically and horizontally and rotating around
the normal to datum S. The tolerance zones
shall remain perpendicular to datum S

Figure 5–34:  Explanation drawing for figure 5–33

188 Find your way in GPS

5.10 Summary
In last chapter we covered the meaning of the 14 tolerance symbols and
looked at examples of how they are used. We covered how some of them
require reference to datums, but did not cover how datums are derived from
datum features and how datum systems work.
In this chapter we have covered the details of the rules for how datums are
derived from datum features. We have established that a datum is always a
plane, a straight line or a point, and that a datum is a mathematically perfect
theoretical feature that does not exist in the real world, as opposed to the da-
tum feature, which is a real surface on the component.
We have looked at the rules for how datums are derived from plane and cy-
lindrical datum features, which are by far the most common. We have also
seen how a conical or wedge shaped datum feature can create more than one
datum.
We have seen that unless otherwise indicated in the drawing, each datum
indicator indicates an entire feature on the component. We have discussed
how you can indicate that only a part of a feature shall be used as the datum
feature, and we have seen how you can use datum targets to indicate that sev-
eral areas of one or more features shall be used together as one datum feature.
We have also covered how you use a common datum to indicate that several
features shall be used simultaneously to create one or more datums.
We have looked at datum systems, where two or three datums are used in a
sequence to create an unambiguous alignment of the component, which al-
lows us to lock the tolerances for all the features in the component together in
the same coordinate system.
Finally, we have looked at TEDs, which are theoretically exact dimensions
that are indicated in rectangular boxes and which do not have tolerances. We
have seen that TEDs are used to position tolerance zones relative to datums
and that some TEDs do not need to be indicated.
In the end we discussed TED patterns, which are a special kind of position
tolerance, where the toleranced features are connected by TED dimensions
and each feature is toleranced relative to the pattern as a whole, so the features
are datum features and toleranced features at the same time.

Find your way in GPS 189

190 Find your way in GPS
6 Combination of size and geometrical tolerances

As mentioned earlier, many GPS tolerances are tailored to ensure that com-
ponents can be assembled and that the features of the components end up
in the right position and orientation relative to the rest of the product after
assembly.
It turns out that the requirements for a component that are necessary to en-
sure assembly often depend on a combination of the size of individual fea-
tures and their relative position and orientation.
Consider, for example, the case of a component with two pins that shall fit
into two holes in another component. The pins will only fit into the holes, if
they have the right size; are in the right distance from each other; and point
in the right direction.
You can use three tolerances to accomplish this: A size tolerance for each pin;
a position tolerance for their relative distance; and an orientation tolerance to
ensure that they point in the right direction.
But it turns out that if the size of the pins is towards the low end of the toler-
ance, then the pins can be further away from the ideal distance and still fit
into the holes. Similarly, they can also be further away from the ideal distance,
or closer to the high end of the size tolerance, if they are perfectly parallel,
than if they are not pointing in the same direction.
So we see that the functional requirement for a set of features being part of an
assembly is a combination of several geometrical requirements. GPS contains
a number of tools that allows you to express these combined requirements.
The advantage of using these tools is that they can create a tolerance that al-
lows unused tolerance for one requirement, for example size, to be used for
another requirement, for example position, exactly in line with the functional
requirements.
This transfer of tolerance from one requirement to another means that it is
easier and cheaper to manufacture components with these tolerances, than if
you did not use a combination requirement, but instead split the tolerance be-
tween the various characteristics in a fixed relationship that still would ensure

Find your way in GPS 191

 assembly, even if size, position and orientation were all right at their respec-
tive tolerance limit.

6.1 Maximum material requirement

Figure 6–1 shows a plug and a hole in a plate. If the requirement merely is
that the thin part of the plug shall fit into the hole, then we can take care of it
with a size tolerance with envelope requirement, as shown in figure 6–1. We
discussed the envelope requirement in chapter 2.

Figure 6–1:  Plug and hole with size tolerances and envelope
requirements

Because we have used an envelope requirement and the upper tolerance limit
(10,0 mm) of the plug is equal to the lower tolerance limit for the hole, we
have ensured that the plug will fit into the hole (in real applications we would
typically leave the upper limit of the plug tolerance slightly smaller than the
lower limit of the hole tolerance). What we have not ensured is that the thin
part of the plug is perpendicular to the step in the plug and that the hole is
perpendicular in the plate. So when we assemble the plug into the hole, it can
look like figure 6–2.

192 Find your way in GPS

 Figure 6–2:  Plug and hole toleranced as in figure 6–1 after
assembly

As we can see in figure 6–2, the plug will fit into the hole, but there is no
guarantee that the step in the plug will fit against the end surface of the hole.
We can tolerance the components as shown in figure 6–3, if we want to ensure
that the step will fit against the end surface.

Figure 6–3:  Plug and hole with size tolerances, envelope

requirements and perpendicularity tolerances

In figure 6–3 we have supplemented the size tolerance with a perpendicu-

larity tolerance. We have had to reduce the size tolerances so the resulting
gap will be at least as large as the allowed perpendicularity error, in order to

Find your way in GPS 193

 ensure a sufficient gap between the hole and the plug to make room for the
angular error.

Figure 6–4:  Explanation drawing for figure 6–3

Figure 6–5:  Plug and hole toleranced as in figure 6–3 after

assembly

As we can see in figures 6-4 and 6-5, the tolerancing in figure 6–3 ensures that
the plug will fit loosely enough in the hole to allow the step in the plug to fit
against the end surface of the hole.
The problem is that we have had to reduce the size tolerances by the amounts
we have allowed for the perpendicularity tolerances. This makes the compo-

194 Find your way in GPS

nents more expensive to manufacture. Figure 6–6 shows graphically what the
tolerance in figure 6–3 allows for the plug.

Figure 6–6:  Allowed combinations of diameter and

perpendicularity for the plug according to the tolerances
in figure 6–3

A partial solution to this problem is shown in figure 6–7. The M indication

in the tolerance indicator means that we have applied the maximum material
requirement. This combines the size tolerance and the geometrical tolerance,
so we can transfer the part of the size tolerance that we do not use to the geo-
metrical tolerance. In fact, we should call it a maximum material allowance
instead of requirement, because it is an allowance to transfer tolerance, not a
more restrictive requirement.

Figure 6–7:  Plug and hole with maximum material

requirement

Find your way in GPS 195

 Formally, the maximum material requirement changes the geometrical toler-
ance from a tolerance that applies to the derived feature (the median line) to
a requirement to the surface.
The requirement creates what is called a maximum material virtual condition
(MMVC), which the material of the component is not allowed to exceed. For
a shaft, which is an outside feature, you find the size of the MMVC feature by
taking the maximum material size (MMS) (the upper tolerance limit), in this
case 10 mm – 0,1 mm = 9,9 mm and add the geometrical tolerance, in this
case 0,1 mm. So MMVC = 10,0 mm for the plug.
The plug is not allowed to exceed the MMS size, 9,9 mm, as a 2-point dia-
meter.
We call the other tolerance limit the least material size (LMS). This is simply
the other tolerance limit, in this case 9,7 mm, and the smallest 2-point dia-
meter, perpendicular to the axis, shall be larger than LMS.
For the hole, which is an inside feature, the maximum material limit is the
lower tolerance limit, so MMS is 10 mm + 0,1 mm = 10,1 mm. For an inside
feature you subtract the geometrical tolerance from MMS to find MMVC, so
MMVC is 10,0 mm – 0,1 mm = 10,0 mm.
The hole is still not allowed to be smaller than the MMS size, 10,1 mm, as a
2-point diameter.
LMS for the hole is the upper tolerance limit, 10,4 mm, that applies perpen-
dicular to the hole axis. All 2-point diameters shall be smaller than this value.
That the two MMVCs have the same diameter means that our tolerance cal-
culations are correct, as the components exactly fit together in the worst case
scenario. The requirements are shown in figure 6–8.

196 Find your way in GPS

 Figure 6–8:  Explanation drawing for figure 6–7

These requirements ensure assembly, as shown in figure 6–5, because the two
MMVC cylinders have the same diameter and each is perpendicular to its
datum.
What we have achieved by using the maximum material requirement is that
if the plug is smaller than its MMS diameter, then the perpendicularity may
exceed its tolerance by the amount the diameter is smaller than MMS. If the
hole is larger than its MMS diameter then the perpendicularity may exceed
its tolerance by the amount the diameter is larger than MMS. With the tol-
erancing in figure 6–3, the perpendicularity was never allowed to exceed 0,1
mm.

Find your way in GPS 197

 Figure 6–9:  Allowed combinations of diameter and
perpendicularity for the plug according to the tolerances
in figure 6–7

Figure 6–9 shows graphically what the tolerance in figure 6–7 allows for the
plug. We can see that the acceptable area is much larger than in figure 6–6.
But we can relax the requirements even further and still ensure the assembly
function. With the maximum material requirement we allowed that the part
of the size tolerance not used could be transferred to the geometrical toler-
ance. But we also have a tool that allows us to transfer unused tolerance in the
other direction. This is called reciprocity and is indicated with R , as shown
in figure 6–10.

Figure 6–10:  Plug and hole with maximum material

requirement and reciprocity

The R modifier cancels the requirement that the feature is not allowed to
exceed MMS, as long as the feature does not exceed MMVC, see figure 6–11.

198 Find your way in GPS

 Figure 6–11:  Explanation drawing for figure 6–10 and 6-13

This requirement is the purest expression of the functional requirement that

the plug shall fit into the hole and the step and that the end surface shall fit
against each other when the components are assembled.

Figure 6–12:  Allowed combinations of diameter and per-

pendicularity for the plug according to the tolerances in
figure 6–10 or 6-13

Figure 6–12 shows graphically what the tolerance in figure 6–10 allows for
the plug. We can see that the acceptable area is even larger than in figure 6–9.
We can also express this requirement another way. This is shown in figure
6–13. Here we have used M with a geometrical tolerance of 0. It may look

Find your way in GPS 199

 impossible to have a tolerance of 0, but it means that geometrical tolerance
may use the unused part of the size tolerance. Consequently, figure 6–13 and
6-10 mean exactly the same thing, but there are fewer calculations involved
in getting from figure 6–13 to the meaning of the tolerance in figure 6–11.
Note however, that we have had to adjust the size tolerance in figure 6–13
compared to 6-10 to avoid the calculations.

Figure 6–13:  Plug and hole with maximum material

requirement and a geometrical tolerance of 0

Figure 6–12 also applies to the tolerance for the plug shown in figure 6–13.

6.2 Maximum material requirement on datums

We have now seen how the maximum material requirement works for an
individual feature of size that references a datum that is not a feature of size.
But if the datum is a feature of size, or we have a datum system where one of
the datums is a feature of size, then the maximum material requirement has
even more possibilities.
Figure 6–14 shows a variation of the plug and hole problem. In this example,
the function is that the plug shall fit into both steps of the hole.

200 Find your way in GPS

 Figure 6–14:  Plug and hole with size tolerances, envelope
requirement and a coaxiality tolerance

The coaxiality tolerance is a fixed tolerance. The meaning would have been
the same, had we used the position symbol.
Figure 6–15 shows the meaning of the tolerances. The thick part of the plug
is limited by a 20 mm envelope cylinder. The thin part of the plug is limited
by a 9,9 mm envelope cylinder. The median line of the thin part of the plug
is limited by a coaxiality tolerance zone with a diameter of 0,1 mm, which is
locked to the axis of the thick part of the plug.
The wide part of the hole is limited by a 20 mm envelope cylinder. The narrow
part of the hole is limited by a 10,1 mm envelope cylinder. The median line
of the narrow part of the hole is limited by a coaxiality tolerance zone with a
diameter of 0,1 mm, which is locked to the axis of the wide part of the hole.
The median lines are not straight lines. They are just shown that way in figure
6–15 for simplicity.

Find your way in GPS 201

 Figure 6–15:  Explanation drawing for figure 6–14

The tolerances are chosen so there is a pair of envelope cylinders that neither
component can exceed, see figure 6–16. Consequently, the tolerancing en-
sures that the components can be assembled.

Figure 6–16:  Plug and hole toleranced as in figure 6–14

after assembly

202 Find your way in GPS

Table 6-1 shows combinations of diameters and allowable coaxiality for the
plug with the tolerancing in figure 6–14. As you can see, the allowable coaxi-
ality is independent of the diameters, as long as these are within their respec-
tive tolerances.

Large plug diameter Small plug diameter Allowable coaxiality

19,8 9,7 0,1
19,8 9,9 0,1
19,8 10,0 Not allowed
20,0 9,7 0,1
20,0 9,9 0,1
20,0 10,0 Not allowed
Table 6–1:  Plug diameters and allowable coaxiality
deviation for the tolerancing in figure 6–14

We can use maximum material requirements on these two components, the

same way we did on the two components in figure 6–7, see figure 6–17.

Figure 6–17:  Plug and hole with maximum material

requirement

By following the rules we see that the MMVC cylinder for the thin part of the
plug is 9,9 mm + 0,1 mm = 10,0 mm and the MMVC cylinder for the narrow
part of the hole is 10,1 mm – 0,1 mm = 10,0 mm. Once again the two MMVCs
have the same diameter, which means that our tolerance calculations are cor-
rect. Assuming, of course, that we want them to just exactly fit into each other
without any room to spare.

Find your way in GPS 203

 Figure 6–18:  Explanation drawing for the tolerances for
the thin part of the plug and the hole in figure 6–17

Since the two MMVC cylinders have the same diameter and each is coaxial
with its datum, these requirements ensure assembly as in figure 6–16.
What we have achieved by using maximum material requirement is that if
the thin part of the plug is smaller than its MMS diameter, then the coaxiality
can exceed its tolerance by the amount the diameter is smaller than MMS,
and if the narrow part of the hole is larger than its MMS diameter, then the
coaxiality can exceed its tolerance by the amount the diameter is larger than
MMS. With the tolerancing in figure 6–14 the coaxiality was never allowed
to exceed 0,1 mm.
Table 6-2 shows combinations of diameters and allowable coaxiality for the
plug with the tolerancing in figure 6–17. The differences from table 6-1 are
shown in red. As you can see, the allowable coaxiality is increased when the
diameter of the toleranced feature becomes smaller.

204 Find your way in GPS

 Large plug diameter Small plug diameter Allowable coaxiality
19,8 9,7 0,3
19,8 9,9 0,1
19,8 10,0 Not allowed
20,0 9,7 0,3
20,0 9,9 0,1
20,0 10,0 Not allowed
Table 6–2:  Plug diameters and allowable coaxiality
deviation for the tolerancing in figure 6–17

We get the same advantages compared to figure 6–17 as in figure 6–10, and
figure 6–13 compared to figure 6–7, by using reciprocity, as shown in figure
6–19, or a coaxiality tolerance of 0 as shown in figure 6–20.

Figure 6–19:  Plug and hole with maximum material

requirement and reciprocity

Find your way in GPS 205

 Figure 6–20:  Plug and hole with maximum material
requirement and a geometrical tolerance of 0

The tolerances in figure 6–19 and figure 6–20 are equivalent and both mean
what is shown in figure 6–21. The difference between figure 6–18 and figure
6–21 is that the MMS requirement is cancelled.

Figure 6–21:  Explanation drawing for the tolerances for

the thin part of the plug and the hole in figure 6–19 and
figure 6–20

206 Find your way in GPS

 Large plug diameter Small plug diameter Allowable coaxiality
19,8 9,7 0,3
19,8 9,9 0,1
19,8 10,0 0,0
20,0 9,7 0,3
20,0 9,9 0,1
20,0 10,0 0,0
Table 6–3:  Plug diameters and allowable coaxiality deviation
for the tolerancing in figure 6–19 and figure 6–20

Table 6-3 shows combinations of diameters and allowable coaxiality for the
plug with the tolerancing in figure 6–19 and 6-20. The differences from table
6-2 are shown in red. As you can see, we allow the diameter of the thin part
of the plug to go all the way to the MMVC diameter, 10,0 mm, as long as the
sum of the diameter and the coaxiality does not exceed the MMVC diameter.
So far this example is exactly like the example in chapter 6.1. But we can also
use maximum material requirement on the datums in this example, because
the datum features are features of size. In the following we continue based on
figure 6–20 with the coaxiality tolerance of 0. We could also have continued
based on figure 6–19, where the result would have been the same, or figure
6–17 where we did not get the advantage of the diameter being allowed to go
all the way to MMVC for the plug and the hole. The change in meaning we
get by using a maximum material requirement on the datum is the same in
all cases.

Figure 6–22:  Plug and hole with maximum material

requirement on both the toleranced feature and the datum
feature and a geometrical tolerance of 0
Find your way in GPS 207
 By adding a maximum material requirement to the datum feature, as in figure
6–22, we are signalling that the datum is no longer the axis of the minimum
circumscribed cylinder for the plug and the maximum inscribed cylinder for
the hole. The datum is now the axis of the MMVC feature for the datum fea-
ture. For the plug this is the upper tolerance limit for the datum diameter, and
for the hole it is the lower limit for the datum diameter. So in both cases, the
datum is the axis in a 20 mm cylinder, regardless of how large or small the da-
tum feature is. Just as we do not need a separate envelope requirement ( ),
when we use a maximum material requirement for the toleranced feature, we
do not need it either, when we use a maximum material requirement for the
datum.
Figure 6–23 shows the meaning of the tolerancing in figure 6–22. In each case
the main requirement is that the component has to clear a MMVC cylinder
for the toleranced feature with a diameter of 10,0 mm and a MMVC cylinder
for the datum feature with a diameter of 20,0 mm, the two MMVC cylinders
being coaxial.
This tolerancing ensures assembly of the two components, because the two
MMCV conditions are the same. Of course this requires that we calculate the
tolerances so this becomes the case. It is not something that happens auto-
matically.
The other requirements are that neither the toleranced feature nor the datum
feature is allowed to be smaller than the LMS size for the plug or larger than
the LMS size for the hole. LMS for the thin part of the plug is 9,7 mm and for
the datum feature of the plug LMS is 19,8 mm. For an external feature LMS is
smaller than MMVC. For the narrow part of the hole LMS is 10,4 mm and for
the datum feature of the hole LMS is 20,2 mm. For an internal feature, LMS
is larger than MMVC.

208 Find your way in GPS

 Figure 6–23:  Explanation drawing for the tolerances in
figure 6–22

Note that when you use a maximum material requirement for both the toler-
anced feature and the datum feature, the two become equal and there is no
difference between the requirements for the two features.

Large plug diameter Small plug diameter Allowable coaxiality

19,8 9,7 0,5
19,8 9,9 0,3
19,8 10,0 0,2
20,0 9,7 0,3
20,0 9,9 0,1
20,0 10,0 0,0
Table 6–4:  Plug diameters and allowable coaxiality
deviation for the tolerancing in figure 6–22

Table 6-4 shows combinations of diameters and allowable coaxiality for the
plug with the tolerancing in figure 6–22. The differences from table 6-3 are
shown in red. As you can see, we allow the coaxiality to increase not only by
the amount that the toleranced feature is not using of its diameter tolerance,
but also by the amount that the datum feature is not using of its diameter tol-
erance. If we compare to table 6-1, we can see that we have increased the co-

Find your way in GPS 209

 axiality tolerance by a factor of 5 in some cases, without having compromised
the function of the components.
The last rule that you have to know in the context of maximum material re-
quirements applies when you indicate a form tolerance with an M indication
for a datum feature that is part of a maximum material requirement for an-
other feature, as in figure 6–24.

Figure 6–24:  Plug and hole with maximum material

requirement and a form requirement to the datum feature

The rule for this indication says that the form tolerance for the datum feature
shall be added to the MMVC diameter for the datum feature for an external
datum feature and subtracted from the MMVC diameter for an internal da-
tum feature. So as shown in figure 6–25, the MMVC diameter for the datum
feature of the plug becomes 20,1 mm and the MMVC diameter for the datum
feature of the hole becomes 19,9 mm. In other words, the tolerances in figure
6–24 do not ensure that the two components can be assembled.

210 Find your way in GPS

 Figure 6–25:  Explanation drawing for the tolerances in figure
6–24. As it can be seen, these tolerances do not ensure that the two
components can be assembled

It might seem backwards that you allow larger deviations by adding an ex-
tra tolerance (straightness) to the drawing, but that is the rule. You should
not have taken the straightness tolerance into account when calculating the
MMVC diameter, if the M had not been indicated in the straightness toler-
ance indicator.
We have now covered all the rules for maximum material requirements, so if
you see them on a drawing, you will be able to decode them. If you are toler-
ancing a drawing yourself, you make it as easy as possible for those who have
to read it by using:
•• A geometrical tolerance of 0.
•• An upper tolerance limit of 0 for external features and datum fea-
tures. In other words, the nominal value on the drawing is the upper
limit.
•• A lower tolerance limit of 0 for internal features and datum features.
In other words, the nominal value on the drawing is the lower limit.
•• No form tolerances with M on datum features.

Find your way in GPS 211

 If you follow these rules, the MMVC diameters become equal to the nominal
diameters given on the drawing. This eliminates possibilities for mathemati-
cal errors where you add where you should have subtracted and vice versa.
Of course, this requires that these rules fit your function. If not, you obviously
have to tolerance according to your functional requirements.

6.3 Maximum material requirement for TED patterns

As we saw in the previous example, the toleranced feature and the datum fea-
ture become equal, when we use maximum material requirements for both
of them.
This ought to remind us of TED patterns, where a number of features are
toleranced relative to each other on an equal basis, where they all are datum
features and toleranced features at the same time. When we use maximum
material requirements for TED patterns, it works in exactly the same way as
when we use maximum material requirements for a toleranced feature and its
datum feature. The only difference is that with TED patterns, more than two
features can be involved in the tolerance.

Figure 6–26:  TED pattern with maximum material

requirement

Figure 6–26 is a variation of figure 5–31. The difference is that we use maxi-
mum material requirement. We have also indicated a size tolerance for the
holes, which is necessary in order to calculate MMVC. We had left out the
size tolerance in figure 5–31, so we could focus on the TED pattern.

212 Find your way in GPS

Because the holes are internal features of size, we calculate the MMVC dia-
meter by taking the lower limit of the size tolerance, 10,2 mm, and subtract-
ing the position tolerance, 0,2 mm, so the MMVC cylinders that the holes
have to fit around are 10,0 mm in diameter.
In addition, the 2-point diameter of the holes shall be larger than the MMS
size of 10,2 mm and less than the LMS size of 10,5 mm everywhere.
We could have eliminated the MMS requirement either by using reciprocity
or by adjusting the lower tolerance limit to 10,0 mm and by using a position
tolerance of 0.

Find your way in GPS 213

 ing

Figure 6–27:  Explanation drawing for the maximum

material requirement in figure 6–26

TED patterns with maximum material requirement are used when a number
of features have to fit together simultaneously. If the component in figure 4–6,
that we have toleranced here, should fit together with the component in figure
4–7, then you could ensure that they could be assembled by using a toleran-
cing for the component with the pins similar to the one we have made for the

214 Find your way in GPS

component with the holes in figure 6–26. In this case you have to make sure
that the MMVC cylinders for the holes are larger than or equal to the MMVC
cylinders for the pins.
If you also want to ensure that the plates can touch each other when the com-
ponents are assembled, you have to reference the surfaces that have to contact
each other as datums for the TED pattern, as in figure 5–33. If, in addition,
you want the side surfaces to line up, when the components are assembled,
you have to reference them as datums and indicate the distance to them with
TEDs, as shown in figure 4–8.
Maximum material requirement is a powerful and functionally correct tool,
when features have to fit together. When you combine it with TED patterns, it
becomes a functionally correct tool for expressing when several features have
to fit together simultaneously.

6.4 Least material requirement

Where maximum material requirement is used to ensure that there is no in-
terference in assembly, least material requirement is used to ensure that there
is interference, for example when a washer has to cover a hole; an interfer-
ence fit has to ensure contact everywhere; a minimum wall thickness shall be
ensured; or to ensure that the finished component can be made from a raw,
unfinished blank or casting.
Let us imagine that the component shown in figure 6–28 is semi-finished and
we want to ensure that we can machine it so we get a finished component
with a ¤ 30 mm outside and a ¤ 15 mm inside diameter and where the hole is
coaxial with the outside surface.
The tolerances in figure 6–28 ensure that the outside 2-point diameter is
above 30 mm everywhere and that the inside 2-point diameter is below 15
mm everywhere. What the tolerances do not ensure is that the component is
straight and the hole is coaxial with the outside diameter.
Note, that where we use the envelope requirement ( ), when we want to en-
sure that a feature does not exceed a perfect feature, we do not have a similar
tool to ensure that the feature will exceed a perfect feature, so the envelope
requirement does not help us here.

Find your way in GPS 215

 Figure 6–28:  Unfinished blank for a component that has
to be ¤ 30 mm outside and have a coaxial ¤ 15 mm hole

Figure 6–29 shows the meaning of the specification in figure 6–28. All the
requirements are 2-point diameters. The outside diameter shall be larger than
30,2 mm and smaller than 30,5 mm everywhere. The inside diameter shall be
larger than 14,6 mm and smaller than 14,9 mm everywhere. The problem is
that the specification does not ensure that the outside cylinder and the inside
cylinder are straight, or that the hole is centred.

Figure 6–29:  Explanation drawing for figure 6–28

Figure 6–30 shows two versions of the left picture of the component in figure
6–28. In the left picture we have used a position tolerance with a least mate-
rial requirement for the outside feature, using the inside feature as the datum
feature. In the right picture we have used a position tolerance with a least

216 Find your way in GPS

material requirement for the inside feature, using the outside feature as the
datum feature.
This allows us to see how we calculate least material virtual condition (LMVC)
for both outside features and inside features. We could have used a coaxiality
tolerance instead of the position tolerance. The result would have been the
same.

Figure 6–30:  Tolerancing alternatives for the unfinished

blank from figure 6–28 with least material requirement for
the outside feature (left) and the inside feature (right)

Figure 6–31 shows the LMVC requirement that corresponds to the left pic-
ture in figure 6–30. For an outside feature of size, the diameter of the LMVC
cylinder is calculated by taking the lower limit of the diameter tolerance (30,0
mm + 0,2 mm = 30,2 mm) and subtracting the geometrical tolerance (0,2
mm). This diameter of the LMVC cylinder is thus 30,0 mm. It is coaxial with
datum C, which is the axis of the maximum inscribed cylinder in the hole.
This is the problem, because it is not the maximum inscribed cylinder (which
is outside the material) that we are interested in. Rather, we are interested in
the minimum circumscribed diameter (which is inside the material).
The other problem is that we still do not have a requirement that ensures that
the hole is straight and round, so there can be a lot of material missing and
the component can still pass the requirement.

Find your way in GPS 217

 Figure 6–31:  The LMVC requirement for the left picture in
figure 6–30. This requirement supplements the diameter
requirements shown in figure 6–29

Figure 6–32 shows the LMVC requirement that corresponds to the right pic-
ture in figure 6–30. For an inside feature of size, the diameter of the LMVC
cylinder is calculated by taking the upper limit of the diameter tolerance (15,0
mm - 0,1 mm = 14,9 mm) and adding the geometrical tolerance (0,1 mm).
This diameter of the LMVC cylinder is thus 15,0 mm. It is coaxial with datum
B, which is the axis of the minimum circumscribed cylinder of the outside
feature. This is again a problem, because it is not the minimum circumscribed
cylinder (which is outside the material) that we are interested in. Rather, we
are interested in the maximum inscribed diameter (which is inside the mate-
rial).
We also still have the problem that we do not have a requirement that ensures
that the outside feature is straight and round, so there can be a lot of material
missing and the component can still pass the requirement.

218 Find your way in GPS

 Figure 6–32:  The LMVC requirement for the right picture
in figure 6–30. This requirement supplements the diameter
requirements shown in figure 6–29

As you may have guessed, the solution to the problem is to add a least mate-
rial requirement to the datum feature in each case, as shown in figure 6–33.

Figure 6–33:  The tolerancing alternatives from figure

6–30 with least material requirement added to the datum
features

By adding a least material requirement to the datum in the tolerance indica-

tor, we are changing the datum so it is no longer the axis of an envelope fea-
ture, as in figure 6–31 and 6-32, but the axis of a LMVC feature.

Find your way in GPS 219

 In figure 6–34, where the datum is an inside feature of size, we find the size
of the LMVC cylinder by taking the upper limit of this size tolerance for the
datum feature, i.e. 14,9 mm in this case.
The requirement in the left picture in figure 6–33 thus ensures that a pair of
coaxial cylinders with diameters of 14,9 mm and 30,0 mm, respectively, can
fit into the component, so they simultaneously are completely inside the ma-
terial.

Figure 6–34:  The LMVC requirements for the left picture

in figure 6–33. These requirements supplement the
diameter requirements shown in figure 6–29

In figure 6–35, where the datum is an outside feature of size, we find the size
of the LMVC cylinder by taking the lower limit of this size tolerance for the
datum feature, i.e. 30,2 mm in this case.
The requirement in the right picture in figure 6–33 thus ensures that a pair
of coaxial cylinders with diameters of 15,0 mm and 30,2 mm, respectively,
can fit into the component, so they simultaneously are completely inside the
material.

220 Find your way in GPS

 Figure 6–35:  The LMVC requirements for the right
picture in figure 6–33. These requirements supplement the
diameter requirements shown in figure 6–29

The requirements in figure 6–33 thus ensure that both a component with an
outside diameter of 30,0 mm and a coaxial hole of 15,0 mm can be machined
from the blank. It does not matter whether you choose one or the other of
the two tolerancing approaches, as long as you choose your tolerance values
correctly. In this case you would probably adjust the tolerance values for the
datum feature, unless you wanted to have a machining allowance on one of
the surfaces.
You can use geometrical tolerances of 0 or reciprocity for least material re-
quirements, just as you can for maximum material requirements. It works
exactly the same way, as long as you remember that the MMVC feature is
outside the material and the LMVC feature is inside the material, and keep
track of plus and minus.
You can also use form tolerances for the datum feature. This moves the LMVC
feature further inside the material, where it moved the MMVC feature fur-
ther out of the material.

6.5 Summary
In this chapter we have looked at how we combine size tolerances and geo-
metrical tolerances. We have looked at maximum material requirements and
least material requirements.
As a general rule, maximum material requirements are used to ensure that
components can be assembled. You define a maximum material virtual con-

Find your way in GPS 221

 dition (MMVC) that the feature is not allowed to exceed with a maximum
material requirement. You create a maximum material requirement by writ-
ing M in the second box in the tolerance indicator.
You can specify an MMVC for an individual feature, either with or without a
relation to a datum. If you relate the MMVC to a datum or a datum system,
you can either reference the datum in the normal way, or you can write M
in that box in the tolerance indicator where you reference a datum for which
the datum feature is a feature of size. If you do the latter, it will have the effect
of making the datum feature equal with the toleranced feature in that both
of them have to fit into each their respective MMVC, which are in perfect
geometrical relationship to each other (e.g. coaxial or parallel in a given dis-
tance), simultaneously.
The maximum material requirement allows the unused part of the size toler-
ance to be used for geometrical deviation. By writing R after M in the toler-
ance indicator, you also allow that unused geometrical tolerance can be used
to exceed the size tolerance. Another way of achieving the same effect is to use
a maximum material requirement with a geometrical tolerance of 0.
A maximum material requirement is in reality an allowance to exchange ex-
cess size tolerance to geometrical tolerance (and vice versa, if you allow reci-
procity). Therefore a tolerance, where you have written M will allow more
components to pass the tolerance, than the same tolerance with the same tol-
erance values, but without M , so in general it is cheaper and easier to manu-
facture components when you have given the maximum material allowance.
You can also use maximum material requirements for TED patterns, either
referenced to the pattern itself or to a datum system. You use this type of tol-
erance when there are several features on the component having to fit with a
counterpart simultaneously. Again in this case, the use of maximum material
requirements makes the components cheaper and easier to produce.
Least material requirements are much less common than maximum material
requirements. Where maximum material requirements define MMVC fea-
tures that the material is not allowed to exceed, least material requirements
define a least material virtual condition (LMVC) feature that has to be com-
pletely contained inside the material.
Least material requirements are typically used to ensure that a component
having to cover a hole in another component (e.g. a washer) can cover the
hole completely to ensure that components have a minimum wall thickness
(e.g. in pressure vessels), or to ensure that there is sufficient material in blanks

222 Find your way in GPS

or raw castings to ensure that the finished component can be machined out
of it.
Where you in principle calculate MMVC by taking the maximum material
size limit and moving the limit out of the material by the amount of the geo-
metrical tolerance, you calculate LMVC by taking the least material size limit
and moving the limit into the material by the amount of the geometrical tol-
erance.
Maximum and least material requirements have an undeserved reputation of
being difficult and making the components more expensive. However it is not
particularly difficult, once you have learned the principles in the preceding
paragraph and it is a good investment to learn how to use these tools, because
they can make the manufacturing of components easier and cheaper, without
compromising the function of the components.

Find your way in GPS 223

224 Find your way in GPS
7 Special drawing indications

There are some special drawing indications that do not appear very often, but
when they do appear, they change the meaning of the specification signifi-
cantly.
In this chapter we will look at four modifiers that can either change the mean-
ing of individual requirements (all around, common zone and projected tol-
erance zone) or the default meaning of all requirements in the drawing (ISO
10579 – NR).

7.1 All around

The all around symbol is a small circle that can be added either to a tolerance
indicator, a surface texture symbol, or an edge symbol.
The all around symbol is used in different contexts, when a requirement ap-
plies to the entire contour in the drawing picture to which the symbol points.
We have already seen the all around symbol in chapter 4 in connection with
the surface profile and line profile symbols, which are the symbols where the
all around symbol is most often used for geometrical tolerances.
Figure 7–1 shows examples where the all around symbol is used for surface
texture and surface profile requirements. Both of these requirements normal-
ly only apply to one individual feature.
In both cases the all around symbol means that the requirement instead ap-
plies to all the features that make up the contour to which the symbol points,
i.e. the features that are highlighted in orange in the right picture in figure
7–1.
It is important to note, especially for geometrical tolerances, that the all
around symbol is a shorthand that is used to avoid having to show an indi-
vidual tolerance indicator for each of the toleranced features. In other words,
it creates an individual requirement for each feature, not one requirement for
all the features. So in figure 7–1, the all around symbol creates six individual
tolerance zones, one for each of the features that make up the contour. These
tolerance zones are not tied together and can move individually. If we want

Find your way in GPS 225

 the requirement to apply as one tolerance zone, we will have to use the CZ or
common zone modifier, which we will discuss later in this chapter.

Figure 7–1:  The all around symbol is added to a tolerance

indicator for a geometrical tolerance and a surface texture symbol

The all around symbol can also be used for edge requirements. We will look
closer at edge requirements in chapter 10. Contrary to geometrical tolerances
and surface texture tolerances, edge requirements do not apply to features,
but for the transition between features.
The requirement applies to the entire contour to which the symbol points,
when the all around symbol is added to an edge symbol,
The requirement applies to both edges, if an edge symbol points to a contour
line that exists on both the front side and the back side of the component in
the drawing picture to which the symbol points.
The combination of these two rules means that the edge symbol in figure
7–2 creates requirements for all the edges that are shown in red in the right
picture.

Figure 7–2:  The all around symbol added to an edge symbol

226 Find your way in GPS

7.2 Common zone
The CZ or common zone modifier is used for geometrical tolerances to indi-
cate that a requirement to a set of features shall be interpreted as one require-
ment, rather than a number of individual requirements. In that sense, the
common zone modifier overrules the principle of independency.
The individual features for which the requirement applies can either be iden-
tified by having a leader line going to each of them from the tolerance indi-
cator, or by using the all around symbol. In the former case, the toleranced
features do not have to be contiguous (connected to each other), as long as
their relative location and orientation are defined, e.g. with TEDs.
The two left pictures in figure 7–3 show the same all around surface profile
requirement, except that in the lower picture the CZ modifier is added. As we
can see in the right pictures, without the CZ modifier, the all around modifier
creates 6 unrelated tolerance zones that each can move individually. With the
CZ modifier, the tolerance zones are locked together, so in this case where the
toleranced features are contiguous, the requirement creates one contiguous
tolerance zone.

Figure 7–3:  The difference in meaning of an all around

requirement without and with a CZ modifier

Find your way in GPS 227

 7.3 Projected tolerance zone
Projected tolerance zone is a modifier, a P that is added to a position require-
ment for the axis of a hole. The hole can either be a plain hole or a threaded
hole.
Projected tolerance zone is used when the functional requirement applies to
the part of what is mounted in the hole that sticks out of the hole, typically a
pin or a screw.
Figure 7–4 shows an example of a projected tolerance zone requirement.

Figure 7–4:  Projected tolerance zone requirement

The projected tolerance zone requirement indication consists of several parts.

First, the projected feature is indicated with a long dashed double-dotted line.
Next, the length of the projection is indicated with P and a non-toleranced
dimension. Finally, the P modifier is indicated in the second box in the toler-
ance indicator.

228 Find your way in GPS

 Figure 7–5:  The meaning of the requirement in figure 7–4

Figure 7–5 shows the meaning of the requirement for the projected feature.
As illustrated, the requirement does not apply to the real feature, only to its
extension.
The result is that the requirement becomes a combination of a location re-
quirement and an orientation requirement for the toleranced feature. The re-
quirement means that there are some combinations of location and orienta-
tion, which are acceptable, that would not be acceptable, if the requirement
applied to the real feature. This also means that there are some combinations,
which would be acceptable for the real feature, that are not acceptable, when
the requirement is for the projected feature.

7.4 Flexible components

is a modifier defined in ISO 10579. It enables us to specify a component in
both free and restrained condition.
All requirements apply to the component in its free state, if ISO 10579 – NR
(NR stands for non-rigid) is not indicated in the drawing.
The ISO 10579 – NR indication changes this, so all requirements apply in a
defined restrained condition, which has to be indicated in the drawing.

Find your way in GPS 229

 Figure 7–6:  Requirement for a flexible component

Figure 7–6 shows an example of requirements for a flexible component. First-

ly, ISO 10579 – NR is indicated in the drawing together with a defined re-
strained condition. This indication is usually made in or near the title block.
When this is indicated, all requirements apply in the restrained condition,
unless otherwise indicated. So the fixed related surface profile tolerance of
0,02 mm in figure 7–6 applies to the component in the restrained condition.
The modifier cancels this. You can think of it as means “free state”. So
the unrelated surface profile tolerance of 0,5 mm in figure 7–6 applies to the
component in free state.
It is not required that all features shall have requirements in both free state
and restrained condition when ISO 10579 – NR is indicated in the drawing.
You can define requirements for each feature in either free state, or restrained
condition, or both.

7.5 Summary
In this chapter we have looked at four modifiers that can change either the
meaning of individual requirements or the default meaning of all require-
ments in the drawing.
The all around modifier can be used for geometrical tolerances, surface tex-
ture tolerances and edge tolerances.

230 Find your way in GPS

For geometrical tolerance and surface texture tolerances it means that the
tolerance applies to all the features that make up the contour to which the
symbol points, rather than just the single feature to which the symbol points.
For edge tolerances, not applying to features, but to the transition between
features, it changes the meaning of the tolerance to apply to all the edges that
make up the contour to which the symbol points.
The all around symbol does not change the meaning of the individual toler-
ance. It is merely a shorthand to indicate that the tolerance applies to several
features (or edges) individually.
The CZ modifier changes a set of geometrical tolerances from applying to a
set of features individually to applying to all the features simultaneously, by
locking all the involved tolerance zones together in their nominal relative
location and orientation, as defined by TEDs.
The P modifier changes position tolerances for holes from applying to the
hole itself to applying to the extension of the hole axis. The modifier is used in
the tolerance indicator and applies to one requirement at a time. The modifier
is also used to define the length of the projection, which is necessary to make
the requirement meaningful.
The ISO 10579 – NR indication means that all requirements in the drawing
by default apply in a restrained condition. This condition shall be indicated in
the drawing. If some requirements for the component apply in its free state,
then the modifier shall be indicated in the tolerance indicator for these
requirements.

Find your way in GPS 231

232 Find your way in GPS
8 General tolerances

General tolerances apply to those dimensions and/or geometrical character-

istics that do not have individual tolerances. General tolerances only apply, if
they are indicated in the drawing in or near the title block.
General dimensional tolerances are defined in ISO 2768-1, and general geo-
metrical tolerances are defined in ISO 2768-2.
According to these standards, general tolerances are based on the idea that a
given workshop or a given supplier has a certain accuracy that can be main-
tained without any special care or effort and where no savings can be ob-
tained by increasing the tolerances beyond this level.
So if you, for example, need components where that which does not matter,
i.e. the non-functional features, is accurate enough, if it is within the general
tolerance grade “medium”, then you can write that on the drawing and send
the drawing to a workshop that can maintain tolerance grade “medium” with-
out special effort, and nobody ever needs to think about the non-toleranced
dimensions to which the general tolerances apply.
It would, of course, be nice if the world worked like that, but in reality only
very few workshops and factories know which tolerance grade they can main-
tain without special care and effort. This varies depending on which machine
is used, not only with regard to machine type, as grinders are more accurate
than lathes and mills, but it also varies from machine to machine, if the work-
shop has more than one model lathe.
The designer usually has even less knowledge about the capability of the
workshop or the supplier. Especially in these times with global outsourcing,
the designer may not even know in which country the component will be
made, when he makes the drawing.
So the conditions given in the standards for using general tolerances are rare-
ly fulfilled.
In reality, the only advantage of general tolerances is that they save a little
time for the designer in finishing the drawing and they may make it look a
little simpler and less intimidating. Finally, they give the designer the warm

Find your way in GPS 233

 fuzzy feeling that the drawing is complete, because any tolerances he may
have missed are covered by the general tolerances.
On the other hand, they force the supplier and everybody else who has to
read the drawing, such as quality control, to study the drawing intensely to
find the hidden tolerances.
In practice it also turns out that general tolerances create many problems and
discussions between customers and suppliers. Aside from the time this takes,
it also results in components that are rejected because the supplier did not
find one of the dimensions that were covered by the general tolerances and
components that have to be accepted, but do not work, because a general tol-
erance was larger than the designer thought.
So the small time saving and the nice sense that the drawing is complete come
at a very high price. So the only recommendation, if you are a designer, is to
resist the temptation of using general tolerances.
The following review of general tolerances shall therefore not be seen as an
encouragement to use them, but solely as a help to those who have to read
drawings that contain them.

8.1 General dimensional tolerances

ISO 2768-1 defines general dimensional tolerances. These apply to linear di-
mensions, angular dimensions and broken edges.
ISO 2768-1 defines four general dimensional tolerance classes:
f = fine
m = medium
c = coarse
v = very coarse
The four tolerance classes are comparable to the tolerance grades IT11, IT13,
IT15 and IT17. However the intervals that each tolerance value applies to are
much coarser than for the real IT grades. This also means that the steps in tol-
erance value from one interval to the next are much larger, so a small change
in nominal dimension can lead to a large change in the tolerance. As you can
see from table 8-1, the steps are up to a factor of 2.

234 Find your way in GPS

 Allowable deviation in mm
Tolerance class
Fine Medium Coarse Very
coarse
0,5 mm up to 3 mm ±0,05 ±0,1 ±0,2 -
3 mm up to 6 mm ±0,05 ±0,1 ±0,3 ±0,5
6 mm up to 30 mm ±0,1 ±0,2 ±0,5 ±1
30 mm up to 120 mm ±0,15 ±0,3 ±0,8 ±1,5
120 mm up to 400 mm ±0,2 ±0,5 ±1,2 ±2,5
400 mm up to 1 000 mm ±0,3 ±0,8 ±2 ±4
1 000 mm up to 2 000 mm ±0,5 ±1,2 ±3 ±6
2 000 mm up to 4 000 mm - ±2 ±4 ±8
Table 8–1:  General tolerances for linear dimensions

Table 8-1 shows the general dimensional tolerances. ISO 2768-1 also contains
tables for angular dimensions and broken edges.
As you can see from table 8-1, general tolerances are symmetrical. This means
that as a general rule, they are unsuitable for fits, where the shaft tolerance has
to be below nominal and the hole tolerance has to be above nominal in order
for the components to fit together.
Because general tolerances are based on the idea of the general accuracy of
a workshop, ISO 2768-1 only applies to machined components (grinding,
turning, milling, etc.) and components made out of sheet metal.
You indicate general tolerances according to ISO 2768-1 with an indication
near the title block as shown in figure 8–1.

ISO 2768-m

Figure 8–1:  Indication of general dimensional tolerances

class medium according to ISO 2768-1

Find your way in GPS 235

 General dimensional tolerances apply to linear dimensions that are indicated,
but do not have individual tolerances. This means that they do not apply to
TEDs, which are used for position tolerance zones. They do not apply to di-
mensions given in parenthesis either, as these are for information only.
General angular tolerances apply to indicated angular dimensions without
individual tolerances and “angular dimensions usually not indicated” (ISO
2768-1 clause 1 note 3 b), for example angles that are nominally 90° or the
angle of regular polygons.
With regard to rejection, both ISO 2768-1 for general dimensional tolerances
and ISO 2768-2 for general geometrical tolerances say:
“Unless otherwise stated, workpieces exceeding the general tolerance shall
not lead to automatic rejection provided that the ability of the workpiece to
function is not impaired.”
This places general tolerances in a peculiar grey area legally, where it has to be
proven, or at least shown to be likely, that the component will not function,
before it can be rejected for exceeding a general tolerance.

8.2 General geometrical tolerances

ISO 2768-2 defines general geometrical tolerances. These include the follow-
ing requirements for individual features:
•• Straightness
•• Flatness
•• Roundness
As well as the following requirements for the relationship between features:
•• Parallelism
•• Perpendicularity
•• Symmetry
•• Circular run-out
But not requirements for:
•• Cylindricity
•• Coaxiality
•• Line profile
•• Surface profile

236 Find your way in GPS

 •• Angularity
•• Position
•• Total run-out
ISO 2768-2 defines three general geometrical tolerance classes:
H = fine
K = medium
L = coarse
The general geometrical tolerances are primarily intended for machined com-
ponents, just like the general dimensional tolerances. But they can also be
used for other types of components, if the producer’s capabilities are known,
according to ISO 2768-2.
Except for circular run-out, the general geometrical tolerances depend on the
dimensions of the features.
Just like the general dimensional tolerances, general geometrical tolerances
shall be indicated near the title block, see figure 8–2.

ISO 2768-K

Figure 8–2:  Indication of general geometrical tolerances

class medium according to ISO 2768-2

Finally, it is possible to indicate that both general dimensional tolerances and

general geometrical tolerances apply to the component, see figure 8–3.

Find your way in GPS 237

 ISO 2768-m-K

Figure 8–3:  Indication of both general dimensional

tolerances class medium according to ISO 2768-1 and
general geometrical tolerances class medium according to
ISO 2768-2

8.3 Summary
General tolerances are a tool that lets the designer signal that he has not
thought about all the dimensions in the drawing and that he prefers to let it be
up to the supplier and the quality control department to spend time finding
all the dimensions and geometrical characteristics that these tolerances apply
to and look up the tolerance values in the ISO 2768 standards.
The prerequisite of using general tolerances is that the designer knows the
capabilities of the work shop or the supplier and knows which tolerances can
be maintained without special considerations or effort. This prerequisite is
rarely fulfilled in practice.
General tolerances are indicated in the drawing with a reference to ISO 2768
in or near the title block. General dimensional tolerances come in four class-
es from fine to very coarse, and general geometrical tolerances come in the
classes fine, medium and coarse. You can specify general dimensional toler-
ances, general geometrical tolerances or both.

238 Find your way in GPS

9 Surface texture tolerances

So far we have looked at how we control the overall geometry of the com-
ponent with GPS tolerances for size and geometry. With these tools we can
ensure that components can be assembled and that the features of the compo-
nents have the correct orientation and location after assembly.
But often the function of the component depends not only on the overall geo-
metry, or the macro geometry as it is also called, but also to a large extent on
the surface texture or micro geometry.
There are many examples of this: The function of features that slide or roll
against each other, e.g. pistons and cylinders in combustion engines, or the
components in a rolling bearing, is highly dependent on the micro geometry
that controls the lubrication, the wear resistance, noise and vibration levels,
etc.
The function of features that are going to be bolted together so they seal either
against gas or liquid leaks, e.g. the surfaces on an engine block and a cylinder
head that are facing each other, highly depends on the micro geometry.
The adhesion of paint and other surface treatments also depends on the mi-
cro geometry.
The tension in a press fit, for example where a bearing is pressed on a shaft,
depends on the micro geometry of the features involved.
Finally, there can be purely cosmetic reasons to have requirements for the mi-
cro geometry. In many cases, such as stainless steel appliances, surfaces have
to be uniform and free from scratches or other defects for the user to find the
product acceptable.
These are but a few examples of situations where the surface texture, or the
micro geometry, is important for the function of the component. When you
work in design and manufacturing, you will often find that it is the surface
quality of a few features that has a large effect on the function of the prod-
uct, and that the finishing of these exact surfaces consume a disproportionate
amount of the manufacturing costs for these components.

Find your way in GPS 239

 9.1 What is surface texture?
Traditionally and qualitatively we split the surface texture into roughness,
waviness and form.
We say that the roughness is the result of the interaction of the manufacturing
process with the surface, such as the characteristic marks left in the machined
surface by a turning tool bit, a milling head, or grinding wheel. Figure 9–1
shows an example of a roughness profile for a ground surface.

Figure 9–1:  Roughness profile for a ground surface

The profile in figure 9–1 looks very rough compared to what you see when
you look at a ground surface with the naked eye. This is because the vertical
scale is magnified far more than the horizontal scale. So while the length of
the profile may be 10 mm, the height of the profile is maybe only 5 µm, or
0,005 mm. This means that the ratio between the two scales in this case is in
the order of magnitude of 200:1, because the profile is printed here such that
it is approximately 10 times longer than it is tall. We call this the aspect ratio.
An aspect ratio of 200:1 is very typical for a profile that has been printed out
from a profilometer.
If we look closer at the roughness profile in figure 9–1, we will also notice that
all the variations in the surface are very short waved in nature. This means
that the distance between neighbouring peaks and valleys is very short in the
horizontal direction.
The profile in figure 9–2 is a waviness profile. We see the waviness as the result
of e.g. vibration during machining or the release of stresses in the material.
Waviness is often very uniform, as shown in figure 9–2, where the distance
between peaks and valleys is constant and the height of each wave is largely
constant as well.

Figure 9–2:  Waviness profile

240 Find your way in GPS

Just like the roughness profile, the waviness profile is also drawn with a large
aspect ratio. If you look at a reflective, but wavy surface, the waviness can look
much worse than it really is. This is an optical illusion that makes it very easy
to see waviness. If, for example, you look at the fuselage of an unpainted air-
craft, the aluminium plates often look very wavy, but if you measured them,
you would find that the amplitude (height) of the waviness is surprisingly
small.
Finally, we see the form deviation as a result of flexing of the guide ways of
manufacturing machines, or simply as the limitations in the geometrical ac-
curacy of the machines. Figure 9–3 shows an example of a form profile, in this
case a straightness profile.

Figure 9–3:  Form profile

In principle, a surface never looks like figure 9–1, 9-2 or 9-3. These effects are
superimposed on one another and are all present to a greater or lesser degree
in all surfaces.
Figure 9–4 shows the waviness profile superimposed on the form profile.

Figure 9–4:  Waviness profile superimposed on form

profile

Finally, figure 9–5 shows the roughness profile superimposed on the waviness
profile and the form profile.

Find your way in GPS 241

 Figure 9–5:  Roughness profile superimposed on waviness
profile and form profile

Figure 9–6 shows the raw unfiltered profile of the surface. We call this profile
the primary profile. It contains roughness components, waviness components
and form components and this is what a real surface typically looks like when
magnified in the vertical direction.
The large difference in magnification in the two directions means that the
angles in the surface are quite distorted and look much steeper than they
really are. So when you look at a printout from a profilometer you have to be
careful not to draw conclusions based on how rough the surface looks.

Figure 9–6:  Raw, unfiltered primary profile of a surface

that contains roughness, waviness and form elements

The subdivision into roughness, waviness and form based on how the three
components are created in the manufacturing process is, as mentioned above,
qualitative. It is impossible to make a reliable subdivision based on these prin-
ciples that can be used in a specification.

242 Find your way in GPS

Additionally, the function of the component does not care about how the
deviations are created. This is why instead we use filters to define what we
consider roughness, waviness and form in a given surface and to separate
them, when we specify or measure surfaces.
Contrary to size and macro geometry, where there are “natural” definitions
for things like diameters and position deviations, everything in micro geo-
metry is based on man-made definitions that are not necessarily self-evident
the way diameters are.
There are many details in these definitions, such as cut-off length and evalu-
ation length (we will get back to these) where a small change can give a large
change in the meaning of a specification or the value of a measurement result.
There is also a large number of parameters that each describes certain aspects
of a surface and that can be used to specify exactly those aspects of the surface
that are important to the function.

9.2 Filtering
We have seen that surfaces contain roughness, waviness and form. The first
step in specifying and measuring surfaces is to separate these different aspects
of the surface with filters.
Profile filters are characterised by their cut-off length and whether they sup-
press the wavelengths that are longer or shorter than the cut-off length.
The standardised cut-off lengths are 0,002 5 mm, 0,008 mm, 0,025 mm, 0,08
mm, 0,25 mm, 0,8 mm, 2,5 mm, 8 mm, and 25 mm.
A filter can suppress the long waves, so you, for example, can look at the
roughness without influence from the waviness and the form.
Another filter might suppress the short waves, so the waviness can be seen
without influence from the roughness.
For example, waves longer than 0,8 mm can be removed with a 0,8 mm filter.
This way you can isolate the roughness when you either measure the rough-
ness or when you write a roughness specification in a drawing.
You can also use a set of two filters to remove waves longer than 2,5 mm or
shorter than 0,25 to isolate a middle-wave range, i.e. the waviness, again ei-
ther in the specification or the measurement.

Find your way in GPS 243

 It has to be emphasised that these cut-off values are examples that are not
necessarily good choices in specific situations.
We can use the profile shown in figure 9–7 to demonstrate the effect of dif-
ferent cut-off lengths. In all cases the filter is a phase-correct Gaussian filter,
which is by far the most common filter type. This filter type is default accord-
ing to the GPS standards. Note that profilometers more than 10-15 years old
are likely to have other types of filters built in that will give different results.

Figure 9–7:  Unfiltered milled surface

In figure 9–8 and 9–9 we have used an 8 mm filter. Figure 9–8 shows the un-
filtered profile overlaid with the long-wave component, i.e. the waviness and
the form. Figure 9–9 shows the short-wave component, i.e. the roughness.

Figure 9–8:  Waviness profile of the milled surface

using an 8 mm cut-off

Figure 9–9:  Roughness profile of the milled surface

using an 8 mm cut-off

244 Find your way in GPS

The roughness profile in figure 9–9 is the difference between the two profiles
in figure 9–8. Note that the only thing the filter has removed is the general
concavity (hollowness) of the profile.
In figure 9–10 and 9–11 we have used a 2,5 mm filter. This means that we
have moved the crossover between roughness and waviness down to a shorter
wavelength, so we characterise more of the surface structure as waviness and
less as roughness.

Figure 9–10:  Waviness profile of the milled surface

using a 2,5 mm cut-off

Figure 9–11:  Roughness profile of the milled surface

using a 2,5 mm cut-off

If we compare figure 9–8 and 9–10 we can see that the waviness profile fol-
lows the waves in the profile better in figure 9–10. In figure 9–11 we can also
see that much of the waviness has been removed from the roughness profile,
but we can still see residual waviness in figure 9–11.
In figure 9–12 an 9–13 we have used a 0,8 mm filter. This means that we
have moved the crossover between roughness and waviness down to an even
shorter wavelength, so we characterise even more of the surface structure as
waviness and even less as roughness.

Find your way in GPS 245

 Figure 9–12:  Waviness profile of the milled surface
using a 0,8 mm cut-off

Figure 9–13:  Roughness profile of the milled surface

using a 0,8 mm cut-off

If we compare figure 9–10 to 9–12 we can see that the waviness profile follows
the waves in the profile even better in figure 9–12. We can also see in figure
9–13 that the waviness has now been removed from the roughness profile.
Intuitively, this is probably the most suitable cut-off length for this profile and
this surface.
It has to be emphasised that this is only an example and that it is not a good
idea to assume that 0,8 mm will be the correct cut-off value for other surfaces
and profiles, without looking at similar plots of them.
But if we continue to reduce the cut-off value down to 0,25 mm for this profile
we can see what happens in figure 9–14 and 9-15.

Figure 9–14:  Waviness profile of the milled surface

using a 0,25 mm cut-off

246 Find your way in GPS

 Figure 9–15:  Roughness profile of the milled
surface using a 0,25 mm cut-off

If we look at figure 9–14, we can see that the waviness profile has started to
get rough, which means that our filter has begun to characterise part of the
roughness as waviness. It is somewhat harder to see the difference between
figure 9–13 and 9–15, but the profile in figure 9–15 is more uniform, because
some of the long-wave roughness has been filtered out.
If we look at figure 9–9 and figure 9–10 compared to figure 9–14 and 9–15,
we can see how much of a difference the cut-off value can make. This is im-
portant, because when we start looking at the different roughness parameters
and waviness parameters in the next clause, we will have to remember that
they are all defined and calculated from the filtered profile. So the cut-off
value has a very significant influence on the parameter values.
On one hand, this means that if you specify roughness, then you also have to
specify a filter and a cut-off value to be sure that the specification means what
you think it means. There are some rules in the GPS standards, which we will
get back to, that select a cut-off value for roughness, if it is not done explicitly
in the drawing, but it is very dangerous to leave the filter choice up to the
standards. If you want to be in charge of what your specification means, you
have to select the cut-off value yourself.
On the other hand, it also means that you always have to pay attention to the
type of filter you use and what your cut-off value is when you measure rough-
ness and waviness. Without that information your measurement results are
meaningless.
In addition to what we normally refer to as the “cut-off filter” that separates
roughness and waviness, the GPS standards also prescribe a so-called λs-
filter. λ (lambda) is the Greek letter we use for wavelength and s means short.
The λs-filter filters the very short wavelengths, normally shorter than 2,5 µm,
out of the surface profile.

Find your way in GPS 247

 The reason that the standards prescribe the λs-filter is that all profiles are fil-
tered, whether you intend it or not. Physically they are filtered by the radius of
the stylus during the measurement, because the stylus is not infinitely sharp
and thin. This means that the stylus cannot penetrate into very small holes in
the surface. Typical stylus radii are 2-5 µm.
The λs-filter was introduced to have a standardised limitation at the short
end of the wavelength range. By using the λs-filter, you avoid problems with
correlation between different instruments, because they may have different
stylus dimensions.

9.3 Basic parameters

A large number of parameters exists for characterising surface texture. In this
clause we will look at the definitions of the most common parameters. We
will get back to the more exotic parameters later. Generally, there are three
families of parameters:
R = Roughness parameters
W = Waviness parameters
P = Primary profile parameters
The roughness parameters thus start with R and are those parameters that
apply to the short-wave component of the surface.
Waviness parameters start with W and are parameters that apply to the
middle-wave component of the surface. Towards the short end, the waviness
is limited by the filter that suppresses the roughness in the profile. Towards
the long end, the waviness is either limited by another filter, or by the length
of the measured profile.
Primary profile parameters start with P and are parameters that apply to
the unfiltered profile (apart from the λs-filter). If you measure a profile that
covers the entire length of the feature, then the primary profile is equal to a
straightness profile for the feature.
For each parameter one or in some cases more letters follows the initial R, W
or P to indicate how the parameter is calculated. So the Ra parameter is cal-
culated the same way as the Wa and Pa parameters, except they are calculated
on profiles that are filtered differently. In the following we will use the R pa-
rameters as examples, but you should keep in mind that the other parameter
families are calculated similarly.

248 Find your way in GPS

There are two lengths we have to consider in the parameter calculation. The
sampling length is equal to the cut-off length. This is the “building block” in
surface texture specification and measurement. The other length is the evalua-
tion length. The evaluation length consists of a number of sampling lengths. If
nothing else is indicated, the evaluation length consists of 5 sampling lengths.
So if you are using a 0,8 mm filter, then the sampling length is 0,8 mm and
the evaluation length is 4 mm (5 × 0,8 mm). It is possible to indicate that the
evaluation length shall be another number of sampling lengths.
Most parameters are calculated over one sampling length and the final result
is the average of all the sampling lengths. There are, however, a few parame-
ters that are calculated directly over the evaluation length.

Figure 9–16:  Surface profile within a sampling length

Figure 9–16 shows a surface profile within a sampling length. We normally

call the horizontal position in the profile x and, as shown in the figure, we re-
fer to the profile height in point x as Z(x). It is this terminology we are going
to use in the formulae for the parameters.

Average parameters
The average parameters are those parameters where every point in the profile
is part of the calculation of the parameter value.

Find your way in GPS 249

 This means that these parameters have a limited response to individual points
tability, but oninthe
theother
profile, for example
hand it meansan isolated
that talllarge
it takes peak.changes
On one in hand
thethis is good for
the repeatability,
ce to make these parameters react. but on the other hand it means that it takes large changes in
the surface to make these parameters react.

Ra
Ra value, which is the most commonly specified parameter, is the
The Ra value,
metical mean deviation of thewhich
profile.is the most commonly specified parameter, is the arith-
metical mean deviation of the profile.
ormula for Ra is:
The formula for Ra is:

1 �
�� � � |����|��
� �

e is no reason toThere is no reason

memorize to any
this (or memorise
of the this
other(orparameter
any of theformulae).
other parameter
It formulae).
It is only included for completeness.
y included for completeness.

Figure 9–17:  Definition of average parameters

Figure
The 9-17: Definition
meaning of average
of the formula for parameters
Ra can most easily be explained with figure
9–17. The red lines represent the deviation of the profile from the median
line in each point. In a measured profile there are between several hundred

14
250 Find your way in GPS
and several thousand points, so the lines here are only representative. The Ra
value is simply the average length of the red lines.
It makes no difference to the calculated value whether you calculate the Ra
value for each sampling length and take the average over the evaluation length
or you calculate the Ra value directly over the evaluation length.
The Ra value thus describes the “average” roughness of the surface. It reacts
to wide peaks and valleys, but not to slender peaks or narrow valleys, which
often subjectively are considered to add more to the roughness.

Figure 9–18:  Surface turned with new cutting tool (top)

and worn cutting tool (bottom)

One consequence of this is that if you manufacture a component with a turned

surface and use the same carbide tool bit and the same cutting parameters,
then the Ra value will often be higher when the tool bit is new and go down
as the tool bit is worn, because the wear means that the tool bit removes less
material in the bottom of the valleys. If you look at the two surfaces in figure
9–18, you will notice that the bottom surface that has been cut with a worn
tool bit looks much rougher than the top one that is turned with a new tool
bit. The Ra value tells the opposite story, because the Ra value is 18 % lower
for the bottom surface.
Another problem that not only applies to Ra, but also Rq, Rz and Rt, which
we will cover below, is that they cannot distinguish between up and down. So
the two surfaces shown in figure 9–19 have the same Ra, Rq, Rz and Rt value,
but it is clear that they have very different functional characteristics. If you

Find your way in GPS 251

other problem that not only applies to Ra, but also Rq, Rz and Rt, which we
cover below, is that they cannot distinguish between up and down. So the
surfaces shown in figure 9-19 have the same Ra, Rq, Rz and Rt value, but it is
r that they have very different functional characteristics. If you imagine that
imaginetwo
two surfaces represent thatbanisters,
the two surfaces
it wouldrepresent
probablytwo
notbanisters,
be hard toit decide
would probably not
ch one you wouldbe difficult
prefer totoslide
decide which one you would prefer to slide down.
down?

Figure 9–19:  Ra, Rq, Rz and Rt cannot distinguish

between these two surfaces
ure 9-19: Ra, Rq, Rz and Rt cannot distinguish between these two surfaces

Rq
in, the formulaAgain,
is onlythe formula
included foriscompleteness:
only included for completeness:

�
1 �
�� � �� � �����
� �

ou are familiar If
with
youstatistics, youwith
are familiar maystatistics,
recognizeyou
thismay
formula as thethis
recognise standard
formula as the stan-
iation. With regards
dard deviation. With regard to figure 9–17, the formula is
to figure 9-17, the formula means that the Rq value the that the Rq
means
are root of the value
squareis of
thethe length
square of of
root each
theof the read
square lines
of the divided
length by the
of each of the read lines
divided by the number of lines. In practical terms it means that the Rq value
will be close to the Ra value in most cases. Theoretically they can be equal, but
16
normally the Rq value is 10-20 % higher than the Ra value. Mathematically,
the Rq value can never be less than the Ra value. The Rq value is farthest from
the Ra value when the surface has some individual high peaks and deep val-
leys that depart from the general “look” of the surface. The Rq value has the
same weaknesses as described for the Ra value. The Rq value is also, like the
Ra value, insensitive to whether you calculate over each sampling length first
and take the average over the evaluation length or you calculate directly for
the entire evaluation length.

252 Find your way in GPS

Extreme parameters
The extreme parameters are those parameters that describe the extremes, for
example highest peak or deepest valley in the profile.

Figure 9–20:  Definition of Rp, Rv and Rz

The four most commonly used extreme parameters are Rp, Rv, Rz and Rt. Rp
quantifies the height of the peaks in the surface. Rv quantifies the depth of
the valleys in the surface. Rz (zehnpunkthöhe in German, ten-point height)
quantifies the height of the roughness, and Rt quantifies the height of the
highest peak and the depth of the lowest valley.

Rp
Figure 9–20 shows a surface profile over an evaluation length that consists of
five sampling lengths. Rp1 is the height of the highest peak over the median
line in the first sampling length; Rp2 is the height of the highest peak over the
median line in the second sampling length, etc. Rp is the average of Rp1, Rp2,
Rp3, Rp4 and Rp5.

Find your way in GPS 253

 Rv
Similarly, Rv1 is the depth of the deepest valley below the median line in the
first sampling length, etc., and Rv is the average of Rv1, Rv2, Rv3, Rv4 and Rv5
in figure 9–20.
Rp and Rv can be used to distinguish between the two surfaces in figure 9–19.

Rz
Finally, Rz1 is the vertical distance from the highest peak to the deepest valley
in the first sampling length. In other words: Rz1=Rp1+Rv1. Rz is the average
of Rz1, Rz2, Rz3, Rz4 and Rz5 in figure 9–20 and therefore Rz=Rp+Rv.
Rz is called the ten-point height because Rz1, Rz2, Rz3, Rz4 and Rz5 each is
the vertical distance between two points in the surface, and because Rz is the
average of the 5 distances, there are ten points in the profile that are part of
the calculation of Rz.
As we can see, Rp, Rv and Rz are calculated by looking at one sampling length
at a time and afterwards averaging over the evaluation length.

Rt
Rt is different from the other extreme parameters in that it is calculated di-
rectly over the evaluation length. Rt is the largest of Rp1, Rp2, Rp3, Rp4 and
Rp5 plus the largest of Rv1, Rv2, Rv3, Rv4 and Rv5. This means that Rt is the
total height of the profile over the evaluation length. Rt and Rz will have the
same value, if the surface is perfectly uniform. In all other cases, Rt will be
larger than Rz.
The Rt value is only based on two points in the evaluation length (highest
peak and lowest valley). Therefore it will exhibit more variance than the Rz
value that is based on ten points, and the Ra and Rq values that are typically
based on several thousand points.

Comparison of parameters
Ra is as mentioned the most commonly used surface roughness parameter,
but it takes it a long time to react to local changes in the surface, because it is
an average of all the points in the profile. Therefore it is rarely suitable to con-
trol the surface function, when anything more than a rough(!) estimate of the
roughness of the surface is required. The same applies to the Rq parameter.

254 Find your way in GPS

For historical reasons, the Rz value is the most commonly used parameter in
German industry. The Rz value is more sensitive to local changes than the Ra
value, so it is quicker to indicate that something has changed in the surface, if
it is used in process control. This also means that it is more “nervous” and will
have a larger standard deviation, if you compare several measurements on the
same surface, or a series of components in production. So on one hand the Rz
value is better for detecting changes in the surface, but on the other hand it
has a larger variance, which is a disadvantage, if you evaluate your measure-
ment statistically. Whether it is an advantage or a disadvantage is a question
of what you want to get out of your measurements.
Finally, Rt is best used for finding local extremes in the surface (if you include
them in your measurement). Rt will exhibit an even higher variance than Rz,
so for Rt the balance is moved even further towards higher sensitivity and
higher statistical variance.
Because waviness profiles and primary profiles usually do not contain five
sampling lengths, Wt and Pt are very often used to specify and characterise
waviness and form.

9.4 Acceptance rules

Surface texture in general and roughness in particular can vary significantly
from one location to another on a surface that is manufactured with the same
process and that looks uniform with the naked eye.
This means that in theory the more measurements you make on a surface, the
higher the highest measured value will be. This means that you have to make
an infinite number of measurements to ensure that every single profile in the
surface is within a specified limit value. This is not practical.
On the other hand there can be a large difference between two surfaces that
have the same mean value for e.g. Ra. One surface can have more or less the
same value throughout the surface, and the other can vary greatly from one
location to the next. This means that the mean value alone is not good for
specifying the surface texture.
These two problems are the background for the somewhat surprising accep-
tance rules for surface texture. The default rule that applies when nothing else
is specified is given in ISO 4288. It basically says that 16 % of the measured
values are allowed to exceed (for an upper limit) the specified parameter
value. Similarly, 16 % of the measured values are allowed to be smaller than a
specified minimum value.

Find your way in GPS 255

 The odd 16 % value comes from statistics. For a normal distribution, 84 %
(100 % - 16 %) of the values will be less than the mean value plus one standard
deviation. On the other hand, 84 % of the values will also be higher than the
mean value minus one standard deviation.

Figure 9–21:  The mean of an upper limit with and without

a “max” indication

A simple and practical way to administer the 16 % rule is that you can accept
the surface if none of the first three measurements exceed the limit. Other-
wise you have to take three additional measurements, and you can accept the
surface if only one of the first six measurements exceed the limit. After that
it is two out of the first 12 measurements or four out of the first 25 measure-
ments that are allowed to exceed the limit.
If no value is allowed to exceed the specified roughness value, you can write
“max” after the specified parameter as shown in the next clause. This is espe-
cially useful for parameters such as Rt, if you want to ensure that the roughest
spot on the surface is below the limit value, e.g. to avoid cracks and metal
fatigue, which is often initiated in the sharpest notch in the surface, which is
normally also the deepest valley.
If no value is allowed to be lower than the specified value, you can write “min”
after the specified parameter.

256 Find your way in GPS

 Figure 9–22:  The meaning of a lower limit with and
without a “min” indication

In other words, a normal roughness specification means that the mean value
plus (or minus) one standard deviation shall be smaller (or larger) than the
specified value, see figure 9–21 and figure 9–22.
The other surprising rule for surface texture is that scratches and defects in
the surfaces are not covered by the surface texture specifications. Later, we
will look at requirements for surface defects and how these are indicated.
In other words, you have to indicate separately, if scratches or defects in the
surface are not allowed.

9.5 Drawing indications

We have now covered the most important rules for specifications of surface
texture, so let us now look at the drawing indications.
There are three basic symbols as shown in figure 9–23. The symbol with the
open triangle means that it is optional whether the surface is manufactured by
removal of material (i.e. cutting, such as turning, milling or grinding) or not.
The symbol with the closed triangle means that the surface shall be manu-
factured by removal of material, and finally the symbol with the circle in the
open triangle means that material removal is not allowed. In other words, it
shall remain as cast or forged or otherwise formed.

Figure 9–23:  The three basic symbols

The symbols can be used without any further indications to simply set re-
quirements for how the surface is manufactured.

Find your way in GPS 257

 Ra 3,2

Figure 9–24:  Symbol for surface texture requirement for

Ra with default rules

Figure 9–24 shows an extension of the basic symbol. When you have addi-
tional requirements for the surface, you add a horizontal line to the longest
leg of the triangle. If there is only one parameter requirement, you write this
under that line. The requirement is written as the parameter followed by the
tolerance value.
In this case the requirement is that the Ra value shall be less than 3,2 µm.
Because the triangle is closed, the surface has to be manufactured by removal
of material. There is a lot of unwritten information in this symbol. First and
foremost the limits for surface parameters are always given in µm, where
everything else in the drawing is given in mm.
Because it is not explicitly written whether it is an upper or a lower limit,
the tolerance shall be interpreted as an upper limit. In addition, as described
above, the 16 % rule applies since nothing else is indicated. This means that
the requirement is that no more than 16 % of the measured Ra values are al-
lowed to exceed 3,2 µm.
In addition it is understood that surface texture requirements always apply
in the direction resulting in the largest parameter values. So if a surface has
different roughness values in different directions, then all the measurements
shall be taken in the direction giving the highest values, typically perpendicu-
lar to the machining marks.

Filter setting
This brings us to the filter. As described in clause 9.2, the filter has a large
influence on the look of the profile from which the parameter is calculated.
As we can see, there is no explicit requirement for what kind of filter shall be
used, as it is always understood that a phase correct Gaussian filter shall be
used. This filter is defined in ISO 16610-21. Work is under way on several
other filters, but as of this writing the Gaussian filter is the only standardised
filter.
Next we have to determine the cut-off value we have to use. ISO 4288 contains
three tables that specify the cut-off value depending on which parameter is
specified and whether the surface is periodic or not. The tables are given here:

258 Find your way in GPS

 Ra Sampling length Evaluation length
[µm] (Cut-off) [mm]
[mm]
(0,006)<Ra≤0,02 0,08 0,40
0,02<Ra≤0,1 0,25 1,25
0,1<Ra≤2 0,8 4
2<Ra≤10 2,5 12,5
10<Ra≤80 8 40
Table 9–1:  Standard cut-off values for non-periodic
surfaces (e.g. ground surfaces) for Ra, Rq and other
average parameters as well as curve parameters

Rz Sampling length Evaluation length

[µm] (Cut-off) [mm]
[mm]
(0,025)<Rz≤0,1 0,08 0,40
0,1<Rz≤0,5 0,25 1,25
0,5<Rz≤10 0,8 4
10<Rz≤50 2,5 12,5
50<Rz≤200 8 40
Table 9–2:  Standard cut-off values for non-periodic
surfaces (e.g. ground surfaces) for Rz, Rp, Rv, Rt and other
extreme parameters

Rsm Sampling length Evaluation length

[mm] (Cut-off) [mm]
[mm]
(0,013)<Rsm≤0,04 0,08 0,40
0,04<Rsm≤0,13 0,25 1,25
0,13<Rsm≤0,4 0,8 4
0,4<Rsm≤1,3 2,5 12,5
1,3<Rsm≤4 8 40
Table 9–3:  Standard cut-off values for periodic surfaces
(e.g. turned surfaces) for all R parameters

Find your way in GPS 259

 Table 9–1, 9–2 and 9–3 are table 1, 2 and 3 from ISO 4288:1996.
As we can see, the first thing we have to determine is whether the surface is
periodic or not. If it is periodic, we have to use table 9–3, where the cut-off
length depends on the Rsm parameter that is the average width of the rough-
ness, i.e. the feed rate of the process that made the surface. We can see that
the table generally requires that the feed rate shall be less than half the cut-off
length, but larger than half the next lower cut-off value.
If the surface is non-periodic, then we have to use table 9–1 or table 9–2,
depending on whether we are measuring average parameters such as Ra or
extreme parameters such as Rz.
In both cases the procedure is that we first have to guess how rough the sur-
face is and make a measurement with the corresponding filter. Let us say we
guess that the Ra value is 3 µm. In that case we first make a measurement with
a 2,5 mm cut-off filter. If the measurement gives us an Ra value of 2,8 µm,
then the value fits within the interval. Had the value instead been e.g. 1,8 µm,
then we would have to try with the next shorter filter, which is 0,8 mm. Had
the value instead been 12 µm, then we would have to try with the next longer
filter, which is 8 mm.
But we are not finished even if our value was 2,8 µm and fitted into the inter-
val for a 2,5 mm cut-off. ISO 4288 prescribes that once we have found a filter
where the value fits into the interval from table 9–1 or table 9–2, we also have
to make a measurement with the next shorter filter, i.e. 0,8 mm cut-off in our
example.
If the value for that last measurement fits into the interval for the shorter fil-
ter, e.g. if our measurement gives a Ra value of 1,8 µm, then the shorter filter
is the correct one and in our example we would have to use a 0,8 mm filter.
If the value does not fall within the interval for the shorter filter, e.g. if our
new Ra value had been 2,1 µm, then the 2,5 mm cut-off filter would have been
the right one.
There are two important conclusions to this. First, if the filter is not indicated
in the drawing, then it is not the parameter value on the drawing that decides
which filter to use, it is the surface itself. This means that the designer does
not have any influence on which filter will be used, so the surface can be sig-
nificantly different from what was expected.
Secondly, you cannot rely on the person measuring the roughness having
understood this very complicated procedure and applied it correctly .

260 Find your way in GPS

 Cut-off Ra Rq Rz Rp Rv Rt
[mm] [µm] [µm] [µm] [µm] [µm] [µm]
0,08 0,280 0,356 1,421 0,687 0,734 3,410
0,25 0,353 0,445 2,075 1,006 1,069 4,192
0,8 0,430 0,549 2,926 1,511 1,414 4,443
2,5 0,595 0,751 4,078 2,080 1,998 5,721
8 0,925 1,140 5,849 3,019 2,830 6,821
Table 9–4:  Parameter values as a function of the cut-off
value for the surface in figure 9–7 to 9–15

As we can see in table 9–4, the roughness values change significantly as a

function of the cut-off value, so the filter choice has a large effect on the evalu-
ation of the surface. This is why you, if you are the designer, always should
indicate the cut-off value together with your surface texture requirements in
your drawings.
If you, for example, want to specify that the roughness requirement in figure
9–24 applies for a 2,5 mm cut-off filter, you have to indicate it as in figure
9–25.
-2,5/Ra 3,2

Figure 9–25:  Symbol for surface texture requirement for

Ra with 2,5 mm cut-off filter

You read the “-“ sign in front of the cut-off value as “to”. That only this value
is indicated means that the λs filter follows the standard and is set at 1/300th
of the cut-off value, i.e. 0,008 mm in this case. It could also have been written
as in figure 9–26.
0,008-2,5/Ra 3,2

Figure 9–26:  Symbol for surface texture requirement for

Ra with 0,008 mm λs filter and 2,5 mm cut-off filter

Combination of requirements
With the tools shown in figure 9–24 to 9–26 we can indicate requirements for
any parameter, and we can specify the cut-off and λs filter settings. As shown
in figure 9–27 we can indicate more than one requirement in one symbol.

Find your way in GPS 261

 -2,5/Ra 3,2
-0,8/Ra 1,8
-2,5/Rz 15

Figure 9–27:  Symbol for surface texture requirement with

more than one requirement

The symbol in figure 9–27 indicates two different requirements for the Ra
value of the surface with different filter as well as a requirement for the Rz
value.

Requirements for W and P parameters

0,8-25/Wt 10 0,008- /Pt 20

Figure 9–28:  Symbols for surface texture requirement for

waviness and primary profile

Figure 9–28 shows how requirements for W and P parameters are indicated.
In these cases the cut-off values always have to be indicated, because there is
no default in the standards. If both requirements applied to the same surface,
they could have been combined as in figure 9–27.

Upper limit, lower limit and maximum requirements

All the requirements we have looked at so far have been upper limit require-
ments where the 16 % rule has applied. But it is also possible to indicate re-
quirements for the lower limit for the surface texture and to remove the 16 %
rule.
-2,5/U Ra 3,2
-2,5/L Ra 1,8
-2,5/Rtmax 15

Figure 9–29:  Symbol for surface texture requirement with

upper, lower and max limits

You write U for upper limit in front of the parameter value, if you want to
emphasise that it is an upper limit, as shown in figure 9–29. You write L to
indicate that this is a lower limit. In both cases the 16 % rule applies. Finally,
you write “max” in continuation of the parameter to indicate that this is an

262 Find your way in GPS

upper limit where the 16 % rule does not apply, so all measured values have
to be within the indicated limit.
Of course the requirements in figure 9–29 can also be made individually or
without explicit indication of the filters.

Number of sampling lengths

-2,5/Rz3 8 -0,8/Ra1 2 -0,8/Rv1max 4

Figure 9–30:  Symbol for surface texture requirement

with non-standard number of sampling lengths in the
evaluation length

The first symbol in figure 9–30 indicates that the Rz value shall be calculated
over 3 sampling lengths instead of the usual 5, because it says Rz3.
The next symbol indicates that the Ra value shall be calculated over one sam-
pling length because it says Ra1. This could, for example, be because the sur-
face is too short for making a measurement with 5 sampling lengths. Apart
from this, it does not make any difference for Ra how many sampling lengths
are used, except the standard deviation of the values changes, which influ-
ences the 16 % rule.
The last symbol in figure 9–30 indicates that the Rv value shall be calculated
over one sampling length, because it says Rv1. The 16 % rule is disregarded,
because it says “max”. This means that the requirement is that the deepest val-
ley in the surface shall be less than 4 µm. This type of requirement is typically
used for surfaces where metal fatigue and scratches are a problem, since these
phenomena typically depend on the sharpest notch, and by limiting the valley
depth, we limit the notch effect.

Surface lay and pattern

You can indicate the type and direction of the surface pattern with the symbol
for surface texture. The seven possibilities are shown in table 9–5. It is com-
paratively rare to see these requirements.

Find your way in GPS 263

 Symbol Meaning
The surface lay shall be parallel to the plane of projection of the
view in which the symbol is used
The surface lay shall be perpendicular to the plane of projec-
tion of the view in which the symbol is used
The surface lay shall be crossed in two oblique directions rela-
tive to the plane of projection of the view in which the symbol
is used
The surface lay shall be multi-directional, i.e. not in just one
direction
The surface lay shall be approximately circular relative to the
centre of the surface to which the symbol applies
The surface lay shall be approximately radial relative to the cen-
tre of the surface to which the symbol applies
The surface lay shall be particulate, non-directional, or protu-
berant
Table 9–5:  Surface lay symbols and their meanings

The surface lay indications can, of course, be combined with all the other
surface texture requirements, as shown in figure 9–31.
-0,8/Ra 1,8

Figure 9–31:  Symbol for surface texture requirement with

requirement for Ra and the surface lay

Machining allowance
The symbol for surface texture can also be used to indicate a machining al-
lowance, e.g. on process drawings. Figure 9–32 shows an example where the
machining allowance is 2 mm.
-0,8/Rz 9
2
Figure 9–32:  Symbol for surface texture requirement with
requirement for Rz and machining allowance indicated in mm

264 Find your way in GPS

Manufacturing method
The manufacturing method, such as turned, milled or ground, and the sur-
face treatment, such as chrome plated, can be indicated in free text over the
horizontal line in the surface texture symbol.
chrome plated and polished
-0,8/Ra 0,5

Figure 9–33:  Symbol for surface texture requirement with

requirement for Ra and the manufacturing method

General surface tolerances

If the same surface requirements apply to the majority of the surfaces of a
component, you can indicate this requirement near the title block in the
drawing. You either indicate all the other surface texture requirements that
apply in parentheses next to the symbol, as shown in figure 9–34, to make it
easier for the reader of the drawing, or just indicate a basic symbol (the first
symbol in figure 9–23) in parentheses, if there are so many other require-
ments that it does not make sense to indicate all of them.

-0,8/Rz 3,0 -0,25/Ra 0,2 -0,8/Rz 1,0

Figure 9–34:  General surface texture requirement

The indication in figure 9–34 means that all surfaces without individual sur-
face texture requirements have to be machined and have an Rz value of less
than 3,0 µm according to the 16 % rule, using a 0,8 mm cut-off filter. Ad-
ditionally, the indication helps the reader by indicating that there are also
two other surface texture requirements in the drawing. Some surfaces shall
have an Ra value of less than 0,2 µm with a 0,25 mm cut-off filter, and other
surfaces shall have an Rz value of less than 1,0 µm with a 0,8 mm filter. These

Find your way in GPS 265

 latter requirements shall also be indicated directly on the surfaces where they
apply. They are only indicated here as a help to the reader.

9.6 Advanced parameters

Basic parameters such as Ra and Rz are used to control the general roughness
of surfaces, typically in situations where the surface is not allowed to be too
rough, but where there is no specific functional requirement for the surface.
But for the surfaces with a very specific surface function, e.g. engine cylin-
ders, hydraulic cylinders, rolling bearing components and sealing surfaces
designed to keep gas or liquid under pressure, the basic parameters are not
sufficient to control the function.

The material ratio curve

Many of the advanced surface texture parameters are based on the material
ratio curve.
2.000

µm

-2.000
0.000 mm 15.243

2.000

µm

-2.000
0 % 100

Figure 9–35:  Roughness profile and associated material

ratio curve

Figure 9–35 shows a roughness profile and its associated material ratio curve.
The material ratio curve shows for each level in the profile what percentage

266 Find your way in GPS

of the length is material and what percentage is air. So the curve starts at 0 %
at the level of the highest peak in the surface and ends at 100 % at the level of
the deepest valley in the surface.
There are two simple ways to indicate requirements for the material ratio
curve.
One is to require that in a certain depth below the highest peak there shall be
at least X % material. For example that in a depth of 2 µm below the highest
peak there shall be at least 50 % material. The problem with this first method
is that it is based on the highest peak, which makes the method very unstable
from a statistical point of view, because it is based on a single peak that might
be so slender that it disappears when the component is first used.
The other method is to require that in a certain depth below the level with a
certain percentage of material, typically 2 % or 5 %, there shall be at least X %
material. For example that in a depth of 1 µm below the level where there is 2
% material there shall be at least 25 % material. The advantage of the second
method is that it is immune to variations in the highest peak, so statistically
speaking it is much more repeatable and it better expresses how the surface
works.
Two more sophisticated methods have been developed to indicate require-
ments for the shape of the material ratio curve. Both have been developed
to indicate requirements for engine cylinders, but they can be used for other
surfaces as well.

The Rk parameter family

The Rk parameter family consists of 5 parameters. Rk is the core roughness,
i.e. the roughness based on the 40 % in the middle of the material ratio curve.
Rpk is the roughness of the peaks in the surface, and Rvk is the roughness of
the valleys in the surface. These 3 parameters have the unit µm. Each of the
two triangles that defines Rpk and Rvk in figure 9–36 has the same area as the
corresponding hatched area.
Rmr1 is the material ratio at the transition between the peaks and the core of
the profile, and Rmr2 is the material ratio at the transition between the core
and the valleys of the profile. These 2 parameters have the unit %.

Find your way in GPS 267

 Rpk

Rk
Rpk
0 % 100
Rmr1 Rmr2
Figure 9–36:  Rk parameter definitions

The Rk parameter family is based on a three layer model of the surface (peaks,
core and valleys). With these parameters you can specify the roughness of
the peaks, core and valleys separately. In addition, you can specify where the
transition between peaks and core and core and valleys shall be located.

The Rxq parameter family

The Rxq parameter family is based on a two layer model of the surface (pla-
teaus and valleys). There is no parameter named Rxq. In this context the ap-
propriate letter is substituted for x in the name of each parameter.
The Rxq parameters are not defined for all surfaces. A surface shall have a
distinct plateaus-and-valleys character for the Rxq algorithms to yield results.
Figure 9–37 shows an example of such a surface.
1.000

µm

-5.000
0.000 mm 14.000

Figure 9–37:  The Rxq parameters are well suited to characterise

this surface that consists of plateaus and valleys

Rpq is the Rq value for the plateaus, and Rvq is the Rq value of the valleys.
Rmq is the material percentage at the intersection between the plateaus and
the valleys, or in other words, the percentage of the surface that is made up
of the plateaus.

268 Find your way in GPS

As mentioned in 9.3, the Rq parameter is equal to the standard deviation in
statistics. The Rxq parameters are calculated by plotting the material ratio
curve on a probability scale i.e. a material ratio axis that is linear in standard
deviations, see figure 9–38, where the upper axis is in standard deviations and
the lower, non-linear axis is in percent.
-3s -2s -1s 0 +1s +2s +3s
1.000

µm

-5.000
0.1 1 10 30 50 70 90 99 99.9
Figure 9–38:  Probability plot of the material ratio curve
for the surface in figure 9–37

The probability plot shows two linear segments. The slope of the nearly hori-
zontal segment is the Rpq value, and the slope of the nearly vertical segment
is the Rvq value. The Rmq value is the material ratio in percent where the
lines fit to the two segments intersect, see figure 9–39.
-3s -2s -1s 0 +1s +2s +3s
Rpq

1.000

µm
Rvq

-5.000
0.1 1 10 30 50 70 90 Rmq 99 99.9

Figure 9–39:  Calculation of the Rxq parameters

You can use the Rxq parameters to control plateau honed surfaces such as en-
gine cylinder liners. You can also use them to control porous surfaces, where
you use the Rpq parameter to control the roughness of the machining with-
out taking the pores into account.

Find your way in GPS 269

 Valley suppression filter
You primarily use the Rk parameters or the Rxq parameters to control surfac-
es such as the one in figure 9–37. It turns out that the normal Gaussian filter
has problems with such surfaces. When the filter tries to compensate for the
valleys, it creates artificial pushed up areas on both sides of each valley that
are not present in the real surface. This makes the plateau part of the surface
appear rougher than it really is. This is why a special valley suppression filter
has been developed. It is defined in ISO 13565-1. This filter is able to filter
plateau surfaces with much less distortion than the normal Gaussian filter. If
you measure Rk parameters or Rxq parameters, you have to make sure you
are using this filter, even if it is not indicated in the drawing. These parameters
are always to be measured with the ISO 13565-1 filter.

9.7 Surface imperfections

As mentioned above, surface texture requirements do not cover scratches or
other imperfections in the surface. The only standard that covers surface im-
perfections is ISO 8785.
ISO 8785 defines a number of characteristics and parameters for surface de-
fects. Fundamentally, a distinction is made between single and combined sur-
face defects. Single defects only go either into or out of the material relative to
the reference surface, see figures 9–40 and 9–41. Combined defects go both
into and out of the material, see figure 9–42.
S
D

Figure 9–40:  Single inwards surface defect

270 Find your way in GPS

 S

H
Figure 9–41:  Single outwards surface defect

S
H

D
Figure 9–42:  Combined surface defect

ISO 8785 defines a number of parameters from the extent (S), height (H) and
depth (D), as shown in figure 9–40 to 9–42. These parameters are shown in
table 9-6.

Parameter Meaning
SIMe Largest dimension parallel to the
reference surface
Surface defect length
SIMw Largest dimension perpendicular to
SIMe and parallel to the reference
Surface defect width surface
SIMsd Largest depth of a single surface de-
fect measured from and perpendicu-
Single surface defect depth lar to the reference surface
SIMcd Largest depth of a combined surface
defect measured from and perpen-
Combined surface defect depth dicular to the reference surface
SIMsh Largest height of a single surface de-
fect measured from and perpendicu-
Single surface defect height lar to the reference surface
(continues)

Find your way in GPS 271

 Parameter Meaning
SIMch Largest height of a combined surface
defect measured from and perpen-
Combined surface defect height dicular to the reference surface
SIMa The area of a surface defect projected
onto the reference surface
Surface defect area
SIMt The total area of all surface defects
inside a given area, projected onto
Total surface defect area the reference surface
SIMn The total number of surface defects
on the surface
Number of surface defects
SIMn/A The total number of surface defects
within the evaluation area A
Number of surface defects per area
Table 9–6:  Surface defect parameters

Figure 9–43 shows an example of a specification where the largest surface

defect length is limited to 10 mm.
SIMe = 10

Figure 9–43:  Specification that limits the surface defect length

to 10 mm

Figure 9–44 shows an example of a specification where the number of surface

defects is limited to 10 for each 50 mm2.
SIMn/A = 10/50 mm-2

Figure 9–44:  Specification that limits the number of

surface defects to 10 for each 50 mm2

272 Find your way in GPS

Finally, figure 9–45 shows an example of a specification that requires that the
surface is free from defects. This is probably the most useful surface defect
specification.
SIMn = 0

Figure 9–45:  Specification that limits the number of

surface defects to 0, i.e. no surface defects are allowed

Surface defect specifications make sense for cast blanks and raw materials,
but it is often unacceptable to have surface imperfections on finished com-
ponents. But if you want to be certain to avoid surface defects, you have to
indicate this in the drawing e.g. as shown in figure 9–45, since surface texture
tolerances do not cover surface defects.

9.8 Summary
In this chapter we have covered tolerances for surface texture. We have looked
at how surface texture consists of roughness, waviness and form error.
We have also looked at how we use Gaussian filters to separate these three
surface texture components. We can move the intersection point between the
components by varying the cut-off length of the filter.
After that, we looked at the most basic surface texture parameters. We have
seen that there are three parameter families, R, W and P. The R parameters are
used to specify the roughness, W parameters are used to specify the waviness
and the P parameters are used to specify the primary profile, which contains
all three components.
Each parameter is identified by one or more lower case letters, which follow
the capital letter that identifies the parameter family, e.g. Ra, Rz and Rt. Each
parameter can be specified for all three profiles, R, W and P. Ra, Wa and Pa,
for example, are calculated the same way, only on each their profile.
We have discussed the surface texture symbol that is used to indicate surface
texture requirements in the drawing. The symbol can express whether it is
allowed, forbidden or optional to remove material from the surface. It can
express upper, lower and max limits for parameters and which cut-off value
shall be used for each parameter. We have seen that the default is an upper
limit according to the 16 % rule. We have also seen that you can specify the

Find your way in GPS 273

 surface lay, a machining allowance and describe the manufacturing method
or surface treatment in free text.
We have covered how you can indicate a general surface specification that ap-
plies to all surfaces that do not have an individual surface tolerance.
We have discussed the Rk and Rxq parameter families as examples of some of
the more advanced parameters that are used to specify surfaces with compli-
cated functional requirements.
Finally, we have looked at how surface texture tolerances do not cover surface
defects and how we can tolerance surface defects with the tools from ISO
8785.

274 Find your way in GPS

10 Edges

Most of the GPS system focuses on tolerancing of features and their size and
geometry. Edges are a kind of stepchildren. There are limited tools available
for tolerancing of edges, and in many cases it is not well defined exactly what
a specification means.
This does not mean that edges are unimportant and that you do not have
to tolerance them. Rather, it means that it is difficult to make unambiguous
definitions for edges, mostly because an edge is a transition zone between two
features and it is difficult to make a clear definition of where the features end
and the edge begins.
Tolerancing of edges can be necessary to ensure a correct interface between
components, such that there is contact between the features where it is in-
tended that there is contact, without them being stuck on the edge.

Figure 10–1:  Hole/shaft interface where the edges are toleranced

wrong, so the shaft cannot get to the bottom of the hole

Find your way in GPS 275

 Figure 10–2:  Hole/shaft interface where the edges are toleranced
correctly, so the shaft can get to the bottom of the hole

Figure 10–1 shows an example where the outside edges have a small radius
and the inside edges have a large radius. This prevents correct assembly. In
contrast, figure 10–2 shows an example where the outside edges have a large
radius and the inside edges have a small radius, so the components can be as-
sembled as envisaged in the design.
Another reason that can make it necessary to tolerance edges is to ease as-
sembly of the product. In this case chamfering can, for example, help shafts
find their way into holes.

276 Find your way in GPS

 Figure 10–3:  Hole/shaft interface where chamfering is
used to ease assembly

Figure 10–3 shows an example where chamfers are used to increase the “tar-
get zone” where the components will slide into each other. This improves the
targeting during assembly.
It can also be necessary to tolerance the edges for cosmetic reasons to give the
component and the product the right look.
For safety reasons it can also be necessary to tolerance edges to prevent them
from cutting people.
Finally, it can be necessary to tolerance edges to avoid increased friction due
to edge load or to prevent movable components from being blocked.
There are four ways to tolerance edges:
•• Tolerancing as edges of undefined shape according to ISO 13715
•• Tolerancing of radii according to ISO 129-1, optionally with refer-
ence to ISO/TR 16570
•• Tolerancing of chamfers according to ISO 129-1
•• Tolerancing of line or surface profile according to ISO 1101.

Find your way in GPS 277

 10.1 Tolerancing of edges of undefined shape according to
ISO 13715
Tolerancing of edges with the symbol defined in ISO 13715 is used when you
do not care about the shape of the edge, but want to control its extent.
Figure 10–4 shows the basic symbol in its simplest form with only a plus sign
that means that more material is allowed than if the adjacent features had a
sharp transition, or a minus sign that means that less material is allowed than
if the adjacent features had a sharp transition.
+ -

Figure 10–4:  ISO 13715 edge symbols with plus and minus
indication, respectively

Figure 10–5 shows examples of outside edges that all have plus material and
would be acceptable for an edge symbol with a plus indication.

Figure 10–5:  Examples of outside edges that are

acceptable for an edge symbol with a plus indication

Figure 10–6 shows examples of inside edges that all have plus material and
would be acceptable for an edge symbol with a plus indication.

Figure 10–6:  Examples of inside edges that are acceptable

for an edge symbol with a plus indication

Figure 10–7 shows examples of outside edges that all have minus material and
would be acceptable for an edge symbol with a minus indication.

278 Find your way in GPS

 Figure 10–7:  Examples of outside edges that are
acceptable for an edge symbol with a minus indication

Figure 10–8 shows examples of inside edges that all have minus material and
would be acceptable for an edge symbol with a minus indication.

Figure 10–8:  Examples of inside edges that are acceptable

for an edge symbol with a minus indication

The symbols in figure 10–4 only require that there shall be plus or minus ma-
terial, respectively. They do not indicate any requirements for how much plus
or minus material is allowed or required and they do not specify the direction
of any undercut or burr that may be present.
Completely sharp corners as shown in figure 10–9 are acceptable for both the
plus and minus indication in the symbol in figure 10–4.

Figure 10–9:  Examples of sharp edges that are acceptable

for both edge symbols with minus indication and edge
symbols with plus indication

The edge symbol can be used with one or two tolerance values. If it is used
with one value, as shown in figure 10–10, then the other tolerance value is 0.
The sign for the values shall always be indicated.

Find your way in GPS 279

 +0,5 -0,5

Figure 10–10:  Edge symbols with one tolerance value

The meaning of the plus tolerance is shown in figure 10–11. For an inside
edge, the edge has to fall away from the theoretical feature inside the toler-
ance value, and the extra material is not allowed to exceed the green zone.
For an outside edge, the burr is not allowed to exceed the upper and the right
edge of the green zone. It is not defined how early the edge is allowed to start
in the light green zone.

0,5
0,5

0,5

0,5

Figure 10–11:  The meaning of the requirement in an edge

symbol with a plus tolerance value

The meaning of the minus tolerance is shown in figure 10–12. For the inside
edge, the lacking material is not allowed to exceed the green zone. It is not
defined how early the edge is allowed to start falling away from the theoreti-
cal feature.
For the outside edge, the edge has to fall away from the theoretical feature
inside the tolerance value, and the lacking material is not allowed to exceed
the green zone.

280 Find your way in GPS

 0,5

0,5

0,5
0,5
Figure 10–12:  The meaning of the requirement in an edge
symbol with a minus tolerance value

If the tolerance symbol has two tolerance values, the higher value is written
above the lower value as shown in figure 10–13. It is possible that one value is
positive and the other is negative, but usually they will either both be positive
or both be negative.
+0,5 -0,2
+0,2 -0,5

Figure 10–13:  Edge symbols with two tolerance values

The meaning of the symbol with two plus tolerances is shown in figure 10–14.
When the lower tolerance value is positive, it means that there shall be extra
material. For an inside edge, the tolerance values define a distance interval
from the theoretical corner where the extra material shall begin. For an out-
side edge, the tolerance values define a distance interval from the theoretical
corner where the point of the burr shall be found.
0,5
0,5

0,2
0,2

0,2

0,5
0,2
0,5

Figure 10–14:  The meaning of the requirement in an edge

symbol with two plus tolerance values

Find your way in GPS 281

 The meaning of the symbol with two minus tolerances is shown in figure
10–15. When the lower tolerance value is negative, it means that there shall
be lacking material. For an inside edge, the tolerance values define a distance
interval from the theoretical corner where the deepest point of the lacking
material shall be found. For an outside edge, the tolerance values define a
distance interval from the theoretical corner where the material shall start
falling away from the theoretical feature.
0,5
0,2
0,5
0,2

0,2

0,5
0,2

0,5

Figure 10–15:  The meaning of the requirement in an edge

symbol with two minus tolerance values

As you can see from figure 10–14 and figure 10–15, the direction of the edge
is undetermined, when the edge symbol is used as shown in figure 10–13. If
the direction of the burr or the undercut is important, this can be specified
as shown in figures 10–16 to 10–19 by moving the tolerance values, so they
indicate that the extent of the burr shall be either horizontal or vertical in the
drawing picture indicated by the leader line. A specification of the direction
of the edge only makes sense for plus tolerances for outside edges and minus
tolerances for inside edges.
0,5

+0,5 0,2

+0,2

Figure 10–16:  Edge symbol that specifies that the edge

shall have plus material in the horizontal direction

282 Find your way in GPS

 +0,5
+0,2

0,5
0,2
Figure 10–17:  Edge symbol that specifies that the edge
shall have plus material in the vertical direction

0,5

0,2

-0,2
-0,5

Figure 10–18:  Edge symbol that specifies that the edge

shall have minus material in the horizontal direction

-0,2
-0,5
0,2

0,5

Figure 10–19:  Edge symbol that specifies that the edge

shall have minus material in the vertical direction

Finally, it is possible to indicate general tolerances for edges with the ISO
13715 symbol, see figure 10–20.

Find your way in GPS 283

 -0,4 +0,4
-0,2 +0,2

Figure 10–20:  Symbols for general tolerances for edges

The symbols in figure 10–20 from left to right means that outside edges have
a minus tolerance of 0,2-0,4; that inside edges have a plus tolerance of 0,2-0,4;
and that there are other edge requirements specified for some edges in the
drawing.

10.2 Tolerancing of radii according to ISO 129-1

Contrary to tolerancing according to ISO 13715, where the edge is consid-
ered a transition between two features, the edge is seen as a separate feature,
when it is toleranced as a radius according to ISO 129-1, see figure 10–21.
R10±1

Figure 10–21:  Edge toleranced as a radius according to

ISO 129-1

In this case the edge is a separate feature that is toleranced with a dimensional
tolerance. As we saw in chapter 2, dimensional tolerances do not lock the
features or their dimensions in a coordinate system. This means that the edge
feature is not locked relative to the two features adjacent to it, and that this
requirement does not ensure a smooth transition, see figure 10–22.

284 Find your way in GPS

 R10 R10 R10

Figure 10–22:  Edges that meet the tolerance in figure

10–21, because the radius tolerance does not control the
centre of the edge feature

Dimensional tolerances do not control the form deviation of the feature

either, see figure 10–23. Finally, dimensions that are not sizes of features of
size are generally not well defined. Even though the radius feature is part of a
circle or a cylinder, it is not a feature of size, because it is not a complete circle
or cylinder.
R9
R11

Figure 10–23:  A radius tolerance is a dimensional

tolerance and does not control the form deviation of the
edge feature

Some of these limitations can to some extent be alleviated, if you follow ISO/
TR 16570. This is one of those standards that only applies, if it is written ex-
plicitly in the drawing. So you have to write “ISO/TR 16570” near the title
block, if you want to use these rules.
ISO/TR 16570 defines a control criterion for radius tolerances that simulates
the use of radius gauges.
For an outside radius, a circle with a radius equal to the upper tolerance limit
shall be able to touch the toleranced feature in the central region without
touching the outer regions, and a circle with a radius equal to the lower toler-
ance limit shall be able to touch the toleranced feature in outer regions with-
out touching the central region, see figure 10–24.

Find your way in GPS 285

 Rmin
Rmax

Figure 10–24:  The control criterion according to

ISO/TR 16570 for an outside radius

For an inside radius, a circle with a radius equal to the upper tolerance limit
shall be able to touch the toleranced feature in outer regions without touching
the central region, and a circle with a radius equal to the lower tolerance limit
shall be able to touch the toleranced feature in central region without touch-
ing the outer regions, see figure 10–25.
Rmax
Rmin

Figure 10–25:  The control criterion according to

ISO/TR 16570 for an inside radius

10.3 Tolerancing of chamfers according to ISO 129-1

Tolerancing of chamfers according to ISO 129-1 has many of the same weak-
nesses as tolerancing of radii with ISO 129-1 and all the general weaknesses of
dimensional tolerancing of non-features of size, that come from it being dif-
ficult to determine from where and to where the tolerance goes and in what
direction it applies.
A chamfer can generally be described by two linear dimensions, see figure
10–26, or one linear dimension and an angle, see figure 10–27. The chamfer is
toleranced by tolerancing these dimensions. The figures illustrate the uncer-
tainties associated with this type of tolerancing, when you have to determine
whether real components meet the tolerances.

286 Find your way in GPS

 Figure 10–26:  Chamfer defined by two linear dimensions

Figure 10–27:  Chamfer defined by one linear dimension

and one angle

You often see chamfers specified as e.g. 0,5 ± 0,1 × 45°, where the linear di-
mension is toleranced, but the angle is given as a nominal value. The extra
problem with such a specification relative to figure 10–27 is that it is not clear
how close to the 45° the angle has to be. It is impossible for real components
to fulfil such a specification, if you interpret the angle to be exact.

10.4 Tolerancing of edges with geometrical tolerances

Every edge that is functional should be toleranced using geometrical toler-
ances. This not only controls the extent, but also the form of the edge. It also
makes it possible to control the transition between the edge zone and the
adjacent features, so, for example, discontinuities can be avoided.
When you tolerance edges using geometrical tolerances, you consider the
edge as a separate feature. This feature has to be defined with TEDs, and if
the location and orientation of the edge feature is important, the tolerance
indication has to reference datums, just like geometrical tolerances for any
other feature.

Find your way in GPS 287

 10

10
0,2 A E

Figure 10–28:  Chamfer toleranced with a surface profile

tolerance

Figure 10–28 shows an example of geometrical tolerancing of a chamfer. The

nominal dimensions of the chamfer (and therefore its orientation and loca-
tion) are defined with TEDs. The tolerance indicator references the two adja-
cent features as datums, so the chamfer feature is completely fixated relative
to these features by its tolerance zone.
10

0,2 C B
C
10

R10

Figure 10–29:  Radius toleranced with a surface profile

tolerance

Figure 10–28 shows an example of geometrical tolerancing of a radius as a

separate geometrical feature. The radius and the location of the centre point
are defined with TEDs. The tolerance indicator references the two adjacent
features as datums, so the radius feature is completely fixated relative to these
features by its tolerance zone.

288 Find your way in GPS

10.5 Summary
An edge forms the transition between two features. The edge itself can either
be considered a feature or just an undefined transition.
Many edges have no real function, but some edges can have a significant in-
fluence on the function of the component. It is especially important to toler-
ance these functional edges correctly and unambiguously to ensure the func-
tion of the component.
There are four different ways to tolerance edges. Some of the GPS edge tools
interpret the edge as a separate feature and others do not.
The first method is to use the edge symbol from ISO 13715. This method does
not consider the edge as a separate feature. With the edge symbol you can
specify if there shall be plus material or minus material.
You can do this without numerical values or you can indicate values for how
much plus and/or minus material that is required and allowed.
The rules are not entirely consistent. For outside edges with minus material
and inside edges with plus material, the tolerance value indicates how far
from the theoretical corner the edge is allowed to deviate from the adjacent
feature. For outside edges with plus material (burrs) and inside edges with
minus material (undercuts), the tolerance value indicates the height or depth
of the edge deviation.
Finally, you can also specify the direction of burrs and undercuts with the ISO
13715 symbol.
The second method is to define the edge as a radius and use dimensional tol-
erances to specify the radius. As described in chapter 2, dimensional toler-
ances are not well defined when they are not applied to features of size, and
radii are not features of size. The problem is that it is difficult to define where
the rounding of the corner begins and ends; and if the rounding does not
have perfect geometry (which it in reality never has), then the radius of the
rounding is not well defined.
Another problem is that a dimensional tolerance by definition does not con-
trol the location of the rounding, so even if the rounding has the specified
radius, it may not be located in the theoretically correct location, so it is pos-
sible to have discontinuities and breaks in the transition between the edge
and the adjacent features, even if the rounding has the right radius.

Find your way in GPS 289

 A variation of this method is to specify that the control criterion from ISO/
TR 16570 shall be used. This criterion simulates the use of radius gauges, so
the requirement is better defined, but it still does not control the form of the
edge.
A third method is to define the edge as a chamfer (i.e. a plane or conical sur-
face that forms an angle to the two adjacent features). A chamfer can either
be defined by two linear dimensions or one linear dimension and an angular
dimension. The chamfer is toleranced by tolerancing these dimensions. Since
the chamfer is not a feature of size, you will have the usual problems with the
dimensional tolerances being ambiguous. Additionally, there is the problem
that if the chamfer is toleranced with an angle, the angle is not well defined,
especially not if the chamfer is convex (bulging outward).
The fourth method is to define the edge with TEDs and tolerance it using geo-
metrical tolerances. These can control the form, location and orientation of
the edge, depending on whether you reference datums in the specification.
This means that the continuity can be ensured. The tolerancing of edges with
geometrical tolerances works just like tolerancing of other features with geo-
metrical tolerances.
The first three types of tolerances are used for non-functional edges, where
the purpose of the specification only is to ensure, for example, that you do not
cut yourself on the edge or to make it a little easier to mount the component
in the product. But geometrical tolerances should be used for all functional
edges, since this is the only unambiguous way to tolerance edges.

290 Find your way in GPS

Appendix A A tolerancing example

We have now covered the GPS system and its main rules. You should now
have an understanding of how the GPS system is structured; what the basic
rules are; and how you read and understand GPS symbols and requirements
in a technical drawing.
This is equivalent to having learned a vocabulary and the general grammar
for a new language. It enables you to understand when others speak or write
the new language. It also enables you to create sentences and express yourself
in the new language.
However, at this stage you may not be able to tell a story in this new language.
In other words, we have not covered how you go about tolerancing a drawing.
We will do that briefly in this appendix by covering the IfGPS 8-point toler-
ancing procedure that is a systematic tolerancing procedure developed by Per
Bennich and Henrik Nielsen.
The IfGPS 8-point procedure contains the following points:
1. Establishment of datum systems
2. Tolerancing of features of size
3. Positioning of features with fixed tolerances
4. Mobile tolerances
5. Unrelated form tolerances
6. Combination of tolerances
7. Surface texture tolerances
8. Edge tolerances
In the example we will use a simple design of a shaft that has to fit into a hous-
ing with two bearing surfaces. The design also contains a washer and a nut to
hold the shaft in place. The complete design is shown in figure A–1.

Find your way in GPS 291

 Figure A–1:  The design used in the example

292 Find your way in GPS

The shaft is shown in figure A–2. It has a hub at the front with six threaded
holes where additional components can be attached, and at the back a washer
can be attached, with a nut to hold the shaft in place in the housing.

Figure A–2:  The shaft used in the example

The housing is shown in figure A–3. The housing has three holes with coun-
terbores for bolts, so it can be attached to a foundation.

Find your way in GPS 293

 Figure A–3:  The bearing housing used in the example

We will focus on the shaft and the housing in this example.

294 Find your way in GPS

Because we have to refer to the individual features in the design, we have
to give them names. Normally this is done informally, so it is only within a
company or a department that you have a common understanding of what
is meant when these names are used. Figure A–4 shows the names we will
use for the shaft features, and figure A–5 shows the names we will use for the
housing features.

Figure A–4:  The names used for the shaft features

Figure A–5:  The names used for the housing features

Find your way in GPS 295

 When you are tolerancing a component or the set of components that make
up the product, you have to start by analysing the function of the component,
i.e. how it works together with the other components in the product. That is
an entire field of study in itself that we will not cover here. If we remain in the
language analogy, then this is similar to a creative writing class for authors.
You do not learn the craft of writing a well written novel that tells a good story
just by learning vocabulary and grammar.
You can identify the functional features of the component from a function
and interface analysis. For the components in our example, these are shown
in figures A–6 and A–7.

Figure A–6:  The functional features of the shaft

The shaft has two bearing surfaces that have to fit into the bearings in the
housing. These features are indicated with red in figure A–6.
The back face of the hub and the end face towards the back of the shaft are
used to fixate the shaft in the axial direction. These features are indicated with
blue in figure A–6.
The small diameter beyond the step is used to centre the washer. This feature
is indicated with purple in figure A–6.

296 Find your way in GPS

The front face of the hub, the centring cylinder, the flat surface on the centring
cylinder and the threaded holes are used to centre and attach an additional
component to the shaft. These features are indicated in green in figure A–6.
Additionally, the thread at the rear of the shaft is used for the nut that holds
the components together. We will not go into the details of tolerancing this
thread here.

Figure A–7:  The functional features of the bearing housing

The housing has two bearing surfaces holding the shaft. These features are
indicated with red in figure A–7.
The outer faces of the housing control the axial position of the shaft. These
features are indicated with blue in figure A–7.

Find your way in GPS 297

 The bottom face of the housing and the holes in it are used to position and
orient the housing in the overall construction. These features are indicated
with orange in figure A–7.
We will only tolerance these functional features of the shaft and the housing
in this example.
It must be emphasised that this is a very simple design and that this function-
al analysis is very rudimentary. For a real design with components and func-
tions that are much more complex than these, you should take the time to
make sure that you really have understood the function(s) of the design cor-
rectly. If you have misunderstood the function, this will be reflected in your
tolerancing, which then will be functionally wrong and therefore misleading.
You will also typically have to use smaller tolerances, if your tolerancing is not
based on the functional datums.
Finally, we are only looking at these two components in isolation. When we
tolerance a product, we have to analyse all the components and how they
interface with each other. We also have to ensure that our choice of datum
systems is consistent in all the components, so they are arranged “back to
back” in the product. That will allow the largest possible tolerances while still
ensuring the function of the product.
In addition, there are also many situations where we have to split a com-
mon tolerance between two or more components, e.g. when components are
stacked on top of each other and what we really want to tolerance is the total
height of all the components.
The tolerance values that are chosen in the rest of this appendix are somewhat
arbitrary. They do not guarantee any particular function (except that the shaft
will fit into the bearings and not lock up when the nut is tightened) and will
with great certainty not be correct, if they are copied uncritically for use on
real components. However, the rank order of the tolerance values for the dif-
ferent tolerances to the same feature is correct.
It is always true that a tolerance zone with more unlocked degrees of freedom
shall have a smaller value than one with fewer unlocked degrees of freedom
for it to create an additional requirement for the feature. How much smaller
depends, of course, on the function, so that is another reason not to blindly
copy the values used in the example.

298 Find your way in GPS

A.1 Establishment of datum systems
The first and most important point in the IfGPS 8-point procedure is to iden-
tify and define the global datum system and, if necessary, any local datum
systems for each component.
As part of defining the datum systems, we also add the necessary tolerances
to ensure the integrity of the datum systems.

A.1.1 The global datum systems

The global datum systems shall reflect how the components are positioned in
the overall product.

A.1.1.1 The global datum system for the shaft

The two bearing surfaces are used together to define a common datum, be-
cause they are equal in determining the axis for the shaft. This common da-
tum is the primary datum for the shaft, because these bearing surfaces to-
gether determine the axis orientation of the shaft in the product.
Since the primary datum is a straight line, it locks 4 degrees of freedom; the
two translations perpendicular to the line and the rotations around these di-
rections, see figure A–8.
At the bottom left in the figure it is shown how the common datum would be
referenced at the end of a tolerance indicator. At the bottom right and middle
the degrees of freedom indicator for the primary datum is shown along with
the shaft seen in the same perspective as the degrees of freedom indicator.

Find your way in GPS 299

 Figure A–8:  The primary datum for the shaft locks 4
degrees of freedom

To ensure the integrity of the primary datum, it is necessary to add tolerances

to at least ensure that the two bearing surfaces are coaxial and cylindrical.
These tolerances are shown in figure A–9.

Figure A–9:  Coaxiality and cylindricity tolerances for the

shaft’s bearing surfaces

Note that the leader lines for the coaxiality tolerances as well as the datum in-
dicators are aligned with dimension lines. This is because the toleranced fea-
tures are the median lines and the datums are the axes of the bearing surfaces.
In contrast, the leader lines for the cylindricity tolerances are not aligned with
the dimension lines. This is because the toleranced features here are the inte-
gral features, the bearing surfaces themselves.

300 Find your way in GPS

Figure A–10 shows the meaning of the coaxiality tolerances. Note, that while
the tolerances are separate, they cannot move relative to each other, because
all their non-redundant degrees of freedom are locked to the same datum, so
the tolerance ensures that the bearing surfaces have a common axis.

Figure A–10:  The meaning of the coaxiality tolerances in

figure A–9

Find your way in GPS 301

 Figure A–11 shows the meaning of the cylindricity tolerances.

Figure A–11:  The meaning of the cylindricity tolerances in

figure A–9

The secondary datum will have to control the last translatory degree of free-
dom: the axial motion of the shaft. There are two candidates for secondary
datum: The back side of the hub and the step towards the back end of the
shaft. If the shaft is pulled in one direction or the other while in use, e.g. if it
were a propeller shaft, where the propeller is either always pushing or always
pulling on the shaft, then the feature that stops the axial motion against that
force would be the proper secondary datum. If that is not the case, we would
choose one or the other (or a common datum between the two) based on
other criteria. For the example, we choose to use the back side of the hub as
the secondary datum. We could imagine that the shaft is the shaft of a wind
turbine, where a constant force is pushing against the hub, so that it is always
the interface between the back side of the hub and the corresponding feature
on the bearing housing that determines the shaft’s axial position. Figure A–12
shows the shaft with the primary and secondary datums. The secondary da-
tum only locks one degree of freedom.

302 Find your way in GPS

 Figure A–12:  The secondary datum for the shaft locks 1
additional degree of freedom

We have added a perpendicularity tolerance to the secondary datum referenc-

ing the primary datum to ensure the internal integrity of the datum system.
To ensure the integrity of the secondary datum, we have also added a flatness
tolerance to it, see figure A–13. Note also, that the datum indicator has been
moved to the tolerance indicator. This is an alternative way to attach the da-
tum indicator. It does not change the meaning.

Figure A–13:  Perpendicularity and flatness tolerances for

the back side of the hub

Find your way in GPS 303

 Figure A–14:  The meaning of the perpendicularity
tolerance for the back side of the hub

Figure A–15:  The meaning of the flatness tolerance for the

back side of the hub

Figure A–16 shows the status after the definition of the global datum system
for the shaft.

304 Find your way in GPS

 Figure A–16:  Global datum system for the shaft with
tolerances to ensure the integrity of the datum system

There is one unlocked degree of freedom in the global datum system, the ro-
tation around the shaft axis, as we can see from the degrees of freedom indi-
cator in figure A–12. That is because the shaft has cylindrical symmetry, and it
is only the features locked to the local datum system, which we will discuss in
A.1.2.1, that need to be locked in their angular position. In the normal case,
the global datum system should lock all degrees of freedom, as the global da-
tum system for the housing does. We will discuss that next.

A.1.1.2 The global datum system for the housing

The housing has two interfaces; one where it is attached to the rest of the
product and one where it interfaces with the shaft.
In principle, we could choose either as the global datum system. The argu-
ment for using the attachment to the rest of the product is that this datum
system defines how the housing is aligned when it is assembled. If we choose
this as the global datum system, we will have to make a local datum system
for the surfaces where it interfaces with the shaft.
The arguments for using the shaft interface features as the global datum sys-
tem are that these will be the features with the tightest tolerances, and what
we really want to define is how the shaft is oriented and located in space when
the entire product is assembled.
These kinds of arguments will have to be explored and debated in each design
to find the right global datum system for each component. Often there will be
more than one correct answer. In this case we will choose to use the attach-

Find your way in GPS 305

 ment to the rest of the product as the global datum system. This also allows us
to look at how we attach the local datum system of the shaft interface to the
global datum system.
The primary datum is therefore the bottom surface of the housing. As this is
a plane surface, it locks three degrees of freedom: The translation along the
normal of the plane, and the rotations around the two directions in the plane,
see figure A–17.

Figure A–17:  The primary datum for the housing locks

3 degrees of freedom

To ensure the integrity of the primary datum, it is necessary to add a flatness

tolerance to the datum feature. This tolerance is shown in figure A–18. The
meaning of the tolerance is the same as explained in figure 4–46.

306 Find your way in GPS

 Figure A–18:  Flatness tolerance for the bottom face of the
housing
The secondary datum is a common datum made up of the three bolt holes. A
common datum of a set of parallel holes defines an axis (centred amongst the
holes) and a plane that contains that axis and locks the rotation around it. The
secondary datum locks the two remaining translatory and the one remaining
rotatory degree of freedom, see figure A–19 and A–20.

Find your way in GPS 307

 Figure A–19:  The global datum system for the bearing housing

308 Find your way in GPS

 Figure A–20:  The global datum system consists of primary
datum plane S and two datum features, an axis and a plane,
from common datum H-H

This is an example of a datum system that locks all degrees of freedom with-
out a tertiary datum.
It should be noted that this interface is overdetermined, so the three holes
will “fight” over which one controls the location of the housing, if they are
not in perfect location and orientation relative to each other. This means that
the tolerances of these bolt holes and the corresponding threaded holes in the
adjacent component have to be managed carefully to make sure that the bolts
will fit and not get into a fight.

Find your way in GPS 309

 Figure A–21:  Position and perpendicularity tolerances for
the three mounting holes that form common datum H-H

Figure A–21 shows the tolerances for the median lines of the three mounting
holes. The three holes form a TED pattern, because their relative positions
are indicated with TEDs and 3× is indicated over the tolerance indicators.
This means that the position tolerance locks the three median lines in their
nominal relative positions. The position tolerance zones are perpendicular to
datum S. They are not locked to the sides of the housing, so they can rotate
together as a set. See figure A–22. Later, we will lock the sides to the mounting
hole axes instead. That will lock the relative rotation.

310 Find your way in GPS

 Figure A–22:  The meaning of the position tolerance to the
median lines of the mounting holes

The perpendicularity tolerance zones are not locked together, because orien-
tation tolerances do not respect linear TEDs, but are locked perpendicular to
datum S. The perpendicularity tolerance obviously has to have a lower tol-
erance value to add to the requirement, as the position tolerance zones are
already locked perpendicular to datum S. See figure A–23.

Find your way in GPS 311

 Figure A–23:  The meaning of the perpendicularity
tolerance to the median lines of the mounting holes

If necessary, a straightness tolerance could also have been added to the me-
dian lines. In this case it should have a lower value than the perpendicularity
tolerance to add to the requirement.

A.1.2 Local datum systems

The shaft requires a local datum system for the interface to the components
attached to the front of the shaft using the six threaded holes.
The bearing housing requires a local datum system for the features that inter-
face with the shaft and the three bolt holes which serve as individual datums
for the counterbores.

312 Find your way in GPS

A.1.2.1 The local datum system for the front of the shaft
We imagine that a component with a centre hole that is flat on one side is at-
tached against the front of the hub of the shaft using six bolts that go into the
threaded holes.
The primary datum feature in the local datum system that defines the inter-
face to this component is the front side of the hub. Since the force from the
six bolts holds the component against this feature, that will be the one that
primarily determines the orientation of the component. Therefore the front
side of the hub shall be the primary datum. Because it is a plane surface, the
datum locks the translatory degree of freedom along the axis of the shaft and
the rotatory degrees of freedom around the two directions in the plane. This
is datum H.
The secondary datum feature is the centring cylinder, which centres the ad-
ditional component. This datum is an axis, so it locks the two remaining
translatory degrees of freedom. This is datum N. Finally, the flat on the side
of the centring cylinder is the tertiary datum feature. This datum locks the
remaining degree of freedom, the rotation around the secondary datum. This
is datum C. Figure A–24 shows the local datum system on the front end of
the shaft.

Figure A–24:  The local datum system on the front of the shaft

The primary and secondary datum features for the local datum system have
tolerances that refer back to the primary datum in the global datum system.

Find your way in GPS 313

 Figure A–25 shows the perpendicularity tolerance and the flatness tolerance
for primary datum H. The meaning of these tolerances is identical to the tol-
erances of datum B explained in figures A-14 and A-15.

Figure A–25:  Tolerances for the primary datum

in the local datum system

Figure A–26 shows the position tolerance and the cylindricity tolerance for
secondary datum N. The meaning of the cylindricity tolerance is identical
with the cylindricity tolerance for datums F and R explained in figure A–11.

Figure A–26:  Tolerances for the secondary datum

in the local datum system

314 Find your way in GPS

Figure A–27 shows the meaning of the position tolerance for the median line
of datum feature N.

Figure A–27:  The meaning of the position tolerance for

the median line of the centring cylinder

You may ask why the tolerance is a position tolerance with datum H as the pri-
mary datum and not a coaxiality tolerance. To answer that question, imagine
that datum H is not perfectly perpendicular to datum F-R and ask yourself,
whether you would prefer the centring cylinder to be perpendicular to datum
H or coaxial with datum F-R in that case. As the component that has to be
attached to the front of the shaft interfaces directly with datum H and not
with datum F-R, the answer is that we would prefer the centring cylinder to
be perpendicular to datum H, see figure A–28. Of course, we would also like
the counterpart to be coaxial with the bearings, which is why we have the
perpendicularity tolerance for the front of the hub shown in figure A–25.

Find your way in GPS 315

 Figure A–28:  It is more important for counterpart fit that
the centring cylinder is perpendicular to the hub, than
coaxial with the bearings

The last degree of freedom for the shaft is the rotation around the axis. We did
not lock this degree of freedom in the global datum system. We choose to use
the flat on the centring cylinder to lock this degree of freedom as the tertiary
datum in the local datum system.
Figure A–29 shows the tolerances for the tertiary datum. We have added a
position tolerance referencing the primary and the secondary datums in the
local datum system and a parallelism tolerance referencing the primary da-
tum in the local datum system to ensure the internal integrity of the datum
system. To ensure the integrity of the tertiary datum, we have also added a
flatness tolerance to it.
We have chosen to reference the tolerances to the primary and secondary da-
tums in the local datum system and not the global datum system, because it is
more important that the feature is orientated and located correctly relative to
the primary and secondary datum in the local datum system, than how it is

316 Find your way in GPS

located relative to the global datum system. To decide to which one it should
be referenced, imagine that the two datums, F-R and N, are not aligned. Then
decide which one it would be preferable that the feature is aligned with in that
case, similar to the discussion about the secondary datum in figure A–28.

Figure A–29:  Position, parallelism and flatness tolerances

for the flat on the centring cylinder

Figure A–30 shows the meaning of these tolerances. The three tolerance zones
are each limited by a pair of planes. The only difference between them is the
number of unlocked degrees of freedom and the distance between the planes.
The distance between the planes is 0,5 mm for the position tolerance. The
tolerance zone is locked perpendicular to primary datum plane H and in a
distance of 5 mm from secondary datum axis N. The tolerance zone can ro-
tate around the primary datum axis, but this degree of freedom is imaginary,
since the component is otherwise symmetrical around that axis.
For the perpendicularity tolerance, the distance between the planes is 0,2
mm. Compared to the position tolerance, the distance from the secondary
datum axis N has been unlocked, but the orientation remains locked. This
means that the 0,2 mm perpendicularity tolerance zone can move within the
0,5 mm position tolerance zone, as long as it remains perpendicular to pri-
mary datum plane H.
Finally, the distance between the planes is 0,1 mm for the flatness tolerance.
Compared to the perpendicularity tolerance, the orientation of the tolerance

Find your way in GPS 317

 zone relative to the primary datum plane has been unlocked. This means that
the 0,1 mm flatness tolerance zone can move freely within the 0,2 mm per-
pendicularity tolerance zone and the 0,5 mm position tolerance zone.
The degrees of freedom indicators in figure A–30 illustrate the difference in
mobility amongst the three tolerance zones.

Figure A–30:  The meaning of the tolerances for the flat

on the centring cylinder

The combination of the three tolerances may seem unnecessarily complicated

and cumbersome. The drawing in figure A–29 would look much simpler had
we simply used a position tolerance of 0,1 mm. If the combination of the three
tolerances ensured the function of the shaft, this one tolerance would as well.
However, by using the three tolerances, we are saying that we can live with a
0,5 mm position tolerance, as long as the perpendicularity is within 0,2 mm
and the flatness is within 0,1 mm. So we have been able to increase the posi-
tion tolerance by a factor 5 and the perpendicularity tolerance by a factor 2,
compared to what we would have required, if we only used a 0,1 mm position
tolerance.
This is important, because position tolerances are much harder to maintain
than perpendicularity and flatness tolerances. Influences like temperature
and tool wear tend to influence position much more than orientation (per-
pendicularity) and form (flatness). So by using a combination of tolerances,
we have made the component much easier to manufacture.

318 Find your way in GPS

Of course, this assumes that these are indeed the functional requirements.
Only knowledge about the specific product can tell us if that is true.
In deciding whether one, two, three or even more tolerances are appropriate,
we also have to look at the relationship between the tolerance values we as-
sign. It might be that looking at the deviations we can accept, we decide that
the factor 2 between the perpendicularity and flatness is too small to make it
any easier to manufacture. In that case, we might decide to change the per-
pendicularity tolerance to 0,1 mm and delete the flatness tolerance.
We have now defined a local datum system for the shaft that locks all degrees
of freedom. We have also added tolerances to the datum features that control
their geometry and their relative orientation and location. This ensures the
integrity of the local datum system.

A.1.2.2 The local datum system for the bearing surfaces in the housing
The features that interface with the shaft define a local datum system in the
housing. Mirroring the global datum for the shaft, the two bearing surfaces
are used as datum features for a common primary datum.
Like the primary datum for the global datum system for the shaft, this datum
is a straight line and locks 4 degrees of freedom: the two translations perpen-
dicular to the line and the rotations around these directions.
The secondary datum feature is the feature that interfaces with the secondary
datum feature for the shaft. This locks the last translatory degree of freedom,
the axial motion of the shaft. We discussed the options for the secondary
datum for the shaft above. Here we have to mirror the choice we made above
and select the feature that the shaft’s secondary datum rests against as the
secondary datum feature in the local datum system.
Since the shaft is free to rotate in the bearing housing, there is no need to lock
that degree of freedom in this local datum system. Therefore, there is no ter-
tiary datum in this datum system. Figure A–31 shows the local datum system
for the bearing surfaces in the housing.
For the datum features on the housing I have chosen to use the same datum
letters I used for the datum features on the shaft that they touch. The only dif-
ference is that I have chosen to use two repeated letters to identify each datum
feature. This allows me to keep track of which features rest against each other,
while still allowing me to differentiate between them. It is not a GPS rule that
it has to be done this way. There are many possible systems for keeping track
of datums. You (or your company) should choose a system to follow.

Find your way in GPS 319

 Figure A–31:  The local datum system for the bearing
surfaces in the housing

The tolerance zones for the housing local datum features also mirror those
for the shaft global datum system, so figures A-10 and A-11 also apply to the
tolerances for the primary datum features FF and RR, and figures A-14 and
A-15 also apply to the tolerances for the secondary datum feature BB.

A.1.2.3 The local datums for the counterbores in the housing

Technically, these are not datum systems, but individual datums. Each hole
serves as the datum for its counterbore, because it is more important that
each counterbore is aligned with its respective hole, than they are aligned
with each other or other features in the bearing housing. We have already
identified the holes as datums H. We indicate H-H in the tolerance indica-
tor, when we refer to the set as a common datum in the global datum system,
but just H, when we refer to the individual datum. When we tolerance the
counterbores in A.3.2, we will look at how this is done in a way that avoids
ambiguity.

320 Find your way in GPS

A.1.3 Position functional features with TEDs
Once we have defined the datum systems in the components, we define the
location and orientation of the functional features in the components with
TEDs.
Figure A–32 shows the shaft with all the datums and the necessary TEDs
indicated.

Figure A–32:  The shaft with datums and TEDs indicated

The TEDs for the shaft indicate the thickness of the hub (10); the distance
from the back face of the hub to the end face (80,25); the diameter of the circle
with the threaded holes; (¤30) and the offset of the flat from the centreline
(5). There are other dimensions we could have added as TEDs, e.g. the width
of the bearing surfaces, but in this case we have decided to consider them less
functionally important, so we will tolerance them with ± tolerances. For the
naming convention for the shaft features, see figure A–4.
Figure A–33 shows the housing with all the datums and the necessary TEDs
indicated.

Find your way in GPS 321

 Figure A–33:  The housing with datums and
TEDs indicated

322 Find your way in GPS

The TEDs for the housing indicate the distance between the two front mount-
ing holes (2 × 20); the axial distance between the front and rear mounting
holes (28); the distance from the front outer face to the front mounting holes
(26); the distance from the front outer face to the rear outer face (80); and the
height of the centres of the bearing holes over the bottom surface (45). Like
for the shaft, there are other dimensions we could have added as TEDs, e.g.
the width of the bearing surfaces, but here we have again decided to consider
them less functionally important, so we will tolerance them with ± tolerances.
For the naming convention for the housing features, see figure A–5.

A.2 Tolerancing of features of size

The next point in the tolerancing procedure is to tolerance those features of
size that participate in fits with dimensional tolerances. In general it will make
sense to use E (see chapter 2) because the features are part of fits.

A.2.1 Tolerancing of features of size on the shaft

There are four features of size on the shaft that are going to participate in fits,
the two bearing surfaces, the centring cylinder and the small diameter. In the
case of the bearings we use g6 tolerances, see chapter 2 and ISO 286-1. Both
limits for this tolerance are under the nominal size, ensuring that the shaft
can fit into bearing holes that are the nominal size. For the centring cylinder
and the small diameter we are using f7 tolerances. These are somewhat wider
tolerances than the one we chose for the bearing surfaces, because we do not
want as tight a fit. On a percentage basis, their upper limits are also closer to
the nominal size. However, because of the difference in diameter, the absolute
upper limit deviation is larger.

Find your way in GPS 323

 Figure A–34:  Tolerancing of the features of size on the
shaft that participate in fits

A.2.2 Tolerancing of features of size on the housing

There are eight features of size on the housing that are going to participate in
fits, the two bearing surfaces, the three mounting holes and the three coun-
terbores for the heads of the mounting screws. In the case of the bearings we
use H7 tolerances, see chapter 2 and ISO 286-1. The lower limit of an H toler-
ance is equal to the nominal size. This ensures that the g6 shaft bearing sur-
faces will fit in the holes. For the mounting holes we are using E8 tolerances.
The lower limit is above the nominal size, ensuring that an h cylinder will fit
through it. We use a higher IT grade since these holes do not have the same
accuracy requirements as the bearing surfaces. Finally for the counterbores,
we are using an H11 tolerance, which is quite large, because these bores only
have to provide clearance for screw heads, which are nominally 10 mm.

324 Find your way in GPS

 Figure A–35:  Tolerancing of the features of size on the
housing that participate in fits

A.3 Positioning of features with fixed tolerances

A fixed tolerance is a tolerance where all the non-redundant degrees of free-
dom have been locked by references to datums.

A.3.1 Positioning of features on the shaft with fixed tolerances

Figure A–36 shows the fixed tolerances for the shaft. These include a position
tolerance for the front face of the hub, a position tolerance for the end face
and a coaxiality tolerance for the small diameter, all referencing the global
datum system and a position tolerance for the six threaded holes in the hub
referencing the local datum system.

Find your way in GPS 325

 Figure A–36:  Fixed tolerances for the functional features
of the shaft

Figure A–37 shows the meaning of the position tolerance for the end face.
Note that the tolerance zone is symmetrical around the nominal distance
with half the tolerance on each side.
This position tolerance ensures that the distance between the back side of
the hub and the end face, measured in the direction of the axis of the shaft,
is between 80,15 mm and 80,35 mm. This is the functional distance we are
interested in to make sure the shaft can rotate, when the washer and nut are
mounted at the back end.
The meaning of the position tolerance for the front face of the hub is similar.
It is the same datums that lock the same degrees of freedom. The tolerance
is 1 mm and the TED is 10, so the tolerance ensures that the thickness of the
hub is between 9,5 mm and 10,5 mm. Note that we have already toleranced
the perpendicularity of the front face of the hub to datum F-R and its flatness
with the tolerances given in figure A–25 in the first point of the tolerancing
procedure.

326 Find your way in GPS

 Figure A–37:  The meaning of the position tolerance for
the end face

Figure A–38 shows the meaning of the coaxiality tolerance for the median
line of the small diameter. Note that we have already toleranced the size of
the small diameter in figure A–34 in the second point in the tolerancing pro-
cedure.

Find your way in GPS 327

 Figure A–38:  The meaning of the coaxiality tolerance for
the median line of the small diameter

Figure A–39 shows the meaning of the position tolerance for the median lines
of the 6 threaded holes. The tolerance in figure A–36 applies to all 6 holes
because it says 6× above the tolerance indicator. Because the holes are shown
symmetrically on a circle, there is an implicit TED of 60° between them, and
because the pattern is shown coaxial with datum feature N, there is an im-
plied TED of 0 between the centre of the circle and the axis of the datum.
Finally, because the pattern is shown aligned with datum C, i.e. they are not
shown at an odd angle, there is an implicit TED aligning the pattern rotation-
ally with the datum.

328 Find your way in GPS

 Figure A–39:  The meaning of the position tolerance for
the median lines of the threaded holes

A.3.2 Positioning of features on the housing with fixed tolerances

Figure A–40 shows the fixed tolerances for the housing. These include a posi-
tion tolerance for the bearing surfaces and a position tolerance for the front
outer face, both referencing the global datum system, and a position tolerance
for the rear outer face, referencing the local datum system.
The first two tolerances position the bearing surfaces and the front outer face
with relatively large tolerances to the global datum system. These are the tol-
erances that determine the orientation and location of the shaft in the global
datum system when the product is assembled. We have already locked the
relative location and orientation of these features amongst themselves with
the much tighter tolerances shown in figure A–31.

Find your way in GPS 329

 Figure A–40:  Fixed tolerances for the functional features
of the housing

Figure A–41 shows the meaning of the position tolerance for the front face.
It locks its location in the nominal distance from the mounting holes, per-
pendicular to the bottom face. The global datum system is explained in figure
A–20.

330 Find your way in GPS

 Figure A–41:  The meaning of the position tolerance for
the front face

Figure A–42 shows the meaning of the position tolerance for the bearing sur-
faces. It locks their distance from the bottom face, and orients and locates
them relative to the bottom face and the mounting holes. Formally, there are
two separate tolerance zones, because there is no CZ indication. However,
since the tolerance zones do not have any unlocked non-redundant degrees
of freedom, they function as one. Note that the coaxiality tolerance in figure
A–31 locks the two median lines much tighter to each other than this toler-
ance does. The combination of the two tolerances means that as long as the
median lines are straight and coaxial with each other within 0,01 mm, they
are free to be aligned with the global datum system only within 1 mm.

Find your way in GPS 331

 Figure A–42:  The meaning of the position tolerance for
the median lines of the bearing surfaces

The last tolerance in figure A–40 is the position tolerance to the rear outer
face relative to the local datum system. This tolerance is similar to the one
shown in figure A–37. The difference is that the TED distance is 80 mm for
the housing and 80,25 for the shaft. This means that the distance between the
outer faces, perpendicular to the bearing axis is between 79,9 mm and 80,1
mm, where the length of the shaft is between 80,15 mm and 80,35 mm. In
other words, these two tolerances together ensure that the distance between
the two faces on the shaft is between 0,05 mm and 0,45 mm longer than the
distance between the outer faces on the housing. This ensures that the shaft
can turn when it is mounted in the housing.
There is one more fixed tolerance for the housing. It is shown in figure A–43.
It is a tolerance fixing each counterbore to its hole. Note that where we use

332 Find your way in GPS

H-H in the tolerance indicator in figure A–40 to indicate a common datum
from all three mounting holes, we only use H in the tolerance indicator in fig-
ure A–43 to indicate that in each case it is only one hole we use as the datum
feature.
In addition, figure A–43 shows how an enlarged view can be used to show
feature details and how we can make the tolerance apply to several identical
details by writing N× (in this case 3×) in the identification of the view.

Figure A–43:  Fixed tolerances for the counterbores for the

mounting holes of the housing

For the meaning of the tolerance, see figure 4–15. The only difference is that
in figure 4–15 a common datum is used to define the datum axis, where only
one cylindrical feature is used in this case.

A.4 Mobile related tolerances

A mobile related tolerance is a tolerance where some, but not all non-redun-
dant degrees of freedom are locked by references to datums.

A.4.1 Mobile related tolerances for the shaft

Figure A–44 shows the mobile related tolerances for the shaft. These include a
perpendicularity tolerance for the end face to the primary datum in the global
datum system and a position tolerance for the median lines of the threaded
holes in the hub to the primary datum in the local datum system. The latter is
a pattern tolerance that applies to all the median lines simultaneously.

Find your way in GPS 333

 Figure A–44:  Mobile related tolerances for the shaft

Figure A–45 shows the meaning of the perpendicularity tolerance for the end
face. Compare figure A–45 to figure A–37 to see the difference between the
constraints on the mobile tolerance and the fixed tolerance. Note that the mo-
bile tolerance in figure A–45 has a smaller tolerance than the fixed tolerance.
The mobile tolerance zone is free to move within the fixed tolerance zone, as
long as it remains perpendicular to datum F-R.

Figure A–45:  The meaning of the perpendicularity

tolerance for the end face

Figure A–46 shows the meaning of the position tolerance for the median lines
of the 6 threaded holes. Like the similar tolerance in figure A–36, the toler-
ance in figure A–44 applies to all 6 holes because it says 6× above the toler-

334 Find your way in GPS

ance indicator. Because the holes are shown symmetrically on a circle, there is
an implicit TED of 60° between them. Because this tolerance only references
primary datum H, the holes are allowed to move as a pattern, as long as the
median lines remain in the theoretical position relative to each other and
perpendicular to datum H. That means that the ¤0,1 mm mobile tolerance
zones can move within the fixed ¤0,2 mm tolerance zones, while respecting
those constraints.

Figure A–46:  The meaning of the mobile position

tolerance for the median lines of the threaded holes

A.4.2 Mobile related tolerances for the housing

Figure A–47 shows the only mobile related tolerance for the housing. It is
a perpendicularity tolerance for the rear outer face, referencing the bearing
axis, which is the primary datum in the local datum system.
The front outer face and the bearing surfaces are both datums in the local
datum system and as such they have been toleranced to each other in the first
point in the procedure, see figure A–31. There is no need for mobile related
tolerances to refine their orientation in the global datum system.

Find your way in GPS 335

 Figure A–47:  Mobile related tolerances for the housing

The tolerance in figure A–47 is similar to the one shown in figure A–45.

A.5 Unrelated form tolerances

An unrelated form tolerance is a tolerance where no degrees of freedom are
locked by references to datums.

A.5.1 Unrelated form tolerances for the shaft

Figure A–48 shows all the unrelated form tolerances for the shaft. Most of
these were added earlier to ensure the internal integrity of the global and local
datum systems. The only new ones are the cylindricity tolerance for the small
diameter and the flatness tolerance for the end face. These are underlined in
red.

336 Find your way in GPS

 Figure A–48:  Unrelated form tolerances for the shaft

The meaning of the flatness tolerance is identical to the flatness tolerance for
the rear face of the hub shown in figure A–15, except the width of the toler-
ance is 0,03 mm. The meaning of the cylindricity tolerance is identical to the
cylindricity tolerance for datums F and R explained in figure A–11, except
there is only one toleranced feature, so there is no common zone modifier,
and the tolerance value is 0,05 mm.

A.5.2 Unrelated form tolerances for the housing

Figure A–49 shows all the unrelated form tolerances for the housing. Most of
these were added earlier to ensure the internal integrity of the global and local
datum systems. The only new one is the flatness tolerance for the rear outer
face. It is underlined in red.

Find your way in GPS 337

 Figure A–49:  Unrelated form tolerances for the housing

The meaning of the flatness tolerance is identical to the flatness tolerance for
the rear face of the hub shown in figure A–15.

A.6 Combination of tolerances

The purpose of this point in the procedure is to increase the flexibility by
combining tolerances to make the component simpler and less expensive to
manufacture.
There are two main categories of tolerances that are candidates for combi-
nations. One is combining coaxiality and cylindricity tolerances into total
run-out tolerances or combining concentricity and roundness tolerances into
circular run-out tolerances. The other is to combine size and geometrical tol-
erances for compound fits, usually by using M and R .

338 Find your way in GPS

A.6.1 Combination of geometrical tolerances
Combination of geometrical tolerances is usually relevant for rotating com-
ponents, see 4.4.
In our example there are two candidates for such a combination, the two bear-
ing surfaces on the shaft and in the housing, respectively. Figure A–50 shows
the combination of the tolerances from figure A–9. The modifier CZ in the
total run-out tolerance for the two bearing surfaces means “common zone”
and indicates that the tolerance applies to the two surfaces as if they were one.
This ensures that the two features have the same diameter. However, it does
not lock the diameter to a particular value. So the size can vary within the
range of the size tolerance in figure A–34.

Figure A–50:  Combined geometric tolerance for the shaft

As the tolerance combines the coaxiality and the cylindricity, the two tol-
erance values are added together. However, since the cylindricity tolerance
and the total run-out tolerance both are radial values for the surface and the
coaxiality tolerance is a diameter value for the axis, a coaxiality tolerance al-
lows twice the radial variation on the surface as a cylindricity tolerance or a
total run-out tolerance of the same value. Therefore the coaxiality tolerance is
multiplied by two before the tolerance values are added together. This allows
the part of the tolerance not used for coaxiality to be used for cylindricity
deviation and vice versa. While the tolerance zone has five unlocked degrees
of freedom, it has to fit the two toleranced features simultaneously because of
the CZ modifier, so it ensures that the two toleranced features have a com-
mon axis and that they have similar diameters. Overall, the tolerance controls
the variation in radius over the entire bearing surfaces to a fixed datum axis,
see figure A–51.

Find your way in GPS 339

 Figure A–51:  The meaning of the combined geometric
tolerance for the shaft

Because the bearing surfaces themselves are the datum features for the com-
mon datum that the coaxiality tolerance references, we could have used a
cylindricity tolerance with CZ modifier instead of the total run-out tolerance
with CZ modifier. That is a special case for this component. In the general
case, where the coaxiality tolerance is referencing other features as datum
features, as in figure 4-14, it is necessary to use a total run-out tolerance to
combine the two tolerances. I have used the total run-out tolerance here to
demonstrate the general case.
Similar to figure A–50, figure A–52 shows the combination of the cylindricity
and coaxiality tolerances from figure A–31. The meaning of the tolerance is
similar to the situation shown in figure A–51, except the bearing surfaces are
holes instead of shaft features.

340 Find your way in GPS

 Figure A–52:  Combined geometric tolerance for
the housing

A.6.2 Combination of size and geometrical tolerances

Combination of size and geometrical tolerances are relevant in situations
where multiple features have to participate in a fit simultaneously. While the
bearing surfaces of the shaft and the housing have to fit together, it is not
simply a fit. The shaft has to be able to rotate and presumably there are ad-
ditional requirements for ensure longevity of the shaft/housing interface. The
threaded holes in the hub are candidates for combination of size and geo-
metrical tolerances, since they all have to fit simultaneously, but since they are
threaded holes and there are separate complications with the tolerancing of
thread, we will not cover that here. So the fits we will focus on are the fits for
the mounting holes in the bottom face of the housing and their counterbores.
Figure A–53 shows the revised size and geometrical tolerances for the hous-
ing. The original size tolerances are shown in figure A–35, and the original
geometrical tolerances are shown in figures A-21 and A-43, respectively.

Find your way in GPS 341

 Figure A–53:  Combination of size and geometrical
tolerances for the housing

Figure A–54 shows the maximum material virtual condition, see chapter 6,
for the pattern formed by the three mounting holes, that is created by the
tolerance to the pattern with reference to datum S. The maximum material
virtual condition consists of three perfect ¤5,97 mm cylinders in the nominal
locations relative to each other and perpendicular to datum S. The 5,97 mm
value is found by taking the maximum material size for the holes, i.e. the
lower tolerance limit, 6,02 and subtracting the position tolerance, 0,05 mm.
The pattern is not allowed to violate this virtual condition, so you can think
of it as a theoretical gage that has to fit into the holes in their entire length.
In addition, the two-point diameter of each hole shall remain within the size
tolerance limits of ¤6,02 to ¤6,038.

342 Find your way in GPS

 Figure A–54:  The maximum material virtual condition for
the three mounting holes

The allowable deviations before the combination of tolerances are shown in

table A–1. The tolerances are combined with M on the geometrical tolerance.
The rules for M are covered in chapter 6.

Hole diameter Allowed position deviation

6,02 mm 0,05 mm
6,038 mm 0,05 mm
Table A–1:  Allowable deviations before combination of
tolerances for the three mounting holes

The allowable deviations after the combination of tolerances are shown in

table A–2, where the changed tolerance is shown in red. The allowed position
deviation is increased by 36 % when the hole diameter is at its maximum and
the clearance of 5,97 mm is still ensured.

Hole diameter Allowed position deviation

6,02 mm 0,05 mm
6,038 mm 0,068 mm
Table A–2:  Allowable deviations after combination of
tolerances for the three mounting holes

Find your way in GPS 343

 Figure A–55 shows the maximum material virtual condition for the toler-
ances for the counterbores referencing the mounting holes. Because the toler-
ances are modified with M both for the tolerance value and for datum H, the
maximum material virtual condition consists of two coaxial cylinders, one
¤6,02 mm that the mounting hole shall not violate and another ¤10,9 mm
that the counterbore shall not violate. These values are the maximum material
limits for the two features, i.e. their lower tolerance limits. Because we have
chosen to use a 0- M tolerance instead of using R , no additional calculations
are necessary. For further explanation, see figures 6–20 and 6–21. In addition,
the two-point diameters of the counterbores are not allowed to exceed ¤11,11
mm.

Figure A–55:  The maximum material virtual condition for

the counterbores

The allowable deviations before the combination of tolerances are shown in

table A–3.

Diameter datum H Counterbore diameter Allowable coaxiality

6,02 mm 10,9 mm Not allowed
6,02 mm 11 mm 0,1 mm
6,02 mm 11,11 mm 0,1 mm
6,038 mm 10,9 mm Not allowed
6,038 mm 11 mm 0,1 mm
6,038 mm 11,11 mm 0,1 mm
Table A–3:  Allowable deviations before combination of
tolerances for the three counterbores

344 Find your way in GPS

The allowable deviations after the combination of tolerances are shown in
table A–4, where the changed tolerances are shown in red. The size tolerance
has been increased from 0,11 mm to 0,21 mm, and the allowed coaxiality
deviation is more than doubled from 0,1 mm to 0,228 mm when the datum
hole diameter and the counterbore diameter are both at their maximum and
the same clearance that was created by the original tolerances is still ensured.

Diameter datum H Counterbore diameter Allowable coaxiality

6,02 mm 10,9 mm 0,0 mm
6,02 mm 11 mm 0,1 mm
6,02 mm 11,11 mm 0,21 mm
6,038 mm 10,9 mm 0,018 mm
6,038 mm 11 mm 0,118 mm
6,038 mm 11,11 mm 0,228 mm
Table A–4:  Allowable deviations after combination of
tolerances for the three counterbores

A.7 Surface texture tolerances

For the surface texture tolerances, our main requirements are for the surfaces
that have to slide against each other.

A.7.1 Surface texture tolerances for the shaft

For the shaft, the two bearing surfaces and the back face of the hub are the
ones that slide against other components. The surface texture tolerances for
the shaft are shown in figure A–56. There are three surfaces that have special
surface texture requirements, the two bearing surfaces and the back face of
the hub. We have chosen an upper limit Rz tolerance of 0,2 µm using a 0,25
mm Gaussian filter with the 16 % rule applied for the bearing surfaces and an
upper limit Rz tolerance of 0,5 µm using a 0,25 mm Gaussian filter with the
16 % rule applied for the back face of the hub. In addition, we have indicated
a general surface texture upper limit Rz tolerance of 3 µm using a 0,8 mm
Gaussian filter with the 16 % rule applied and a general requirement that sur-
face defects are not acceptable. See figure 9–34 for an explanation of the gen-
eral surface texture tolerance. Note that a general surface defects specification
can be written near the title block like the general surface texture tolerance.
The SIMn=0 requirement means that no surface defects are acceptable, i.e.
scratches and marks are not allowed on the component.

Find your way in GPS 345

 Figure A–56:  Surface texture tolerances for the shaft

A.7.2 Surface texture tolerances for the housing

For the housing, the two bearing surfaces and the two outer faces are the four
features that slide against other components. The surface texture tolerances
for the housing are shown in figure A–57. We have chosen a fairly complex
set of tolerances for the bearing surfaces. There are three requirements, all of
them from the Rxq family, see clause 9.6. That we are using parameters from
this family means that we want a plateaus and valleys type surface like the one
shown in figure 9–37. The first requirement is an upper limit tolerance of 0,1
µm for Rpq, which limits the roughness of the plateaus. The other two limit
Rvq to be between 5 µm and 8 µm, defining a range for the roughness of the
valleys. All of these requirements apply a 2,5 mm valley suppression filter and
the 16 % rule.
For the outer faces, we have used the same surface texture tolerance we used
for the back face of the hub, an upper limit Rz tolerance of 0,5 µm using a
0,25 mm Gaussian filter with the 16 % rule applied. In addition, we have in-
dicated a general surface texture upper limit Rz tolerance of 3 µm using a 0,8
mm Gaussian filter with the 16 % rule applied and a general requirement that
surface defects are not acceptable, like we did for the shaft.

346 Find your way in GPS

 Figure A–57:  Surface texture tolerances for the housing

A.8 Edge tolerances

The last point in the tolerancing procedure is tolerancing of edges. There
are several ways to tolerance edges, depending on their function, as we saw
in chapter 10. The primary interface where we have to worry about edges
is where the front bearing/back face of the hub on the shaft meets the front
bearing/front outer face of the housing. We need to make sure that the radius
on the housing is greater than the radius on the shaft to ensure that the back
face of the hub on the shaft can rest against the front outer face on the hous-
ing.

A.8.1 Edge tolerances for the shaft

As the shaft presumably is hardened, we worry about stress fractures. One
way to limit these is to ensure that there are no sharp inside corners. For the
transition from the back face of the hub to the front bearing, we specify a ra-

Find your way in GPS 347

 dius using a surface profile tolerance referencing the two adjacent features as
datums. This way we control the geometry of the transition between the two
features completely.

Figure A–58:  Edge tolerances for the shaft

I only show this one tolerance for illustration. For a real component, we would
also have used a similar tolerance for the transition between the end face and
the small diameter; between the front face of the hub and the centring cylin-
der; and between the two bearings and the middle cylinder.
For the edges without tolerances, I have shown a general tolerance that re-
quires outside corners to lack some material and inside corners to have extra
material, see the right picture in figure 10–12 and the left picture in figure
10–14 for an explanation of these general tolerances.

A.8.2 Edge tolerances for the housing

The main edge we are concerned about on the housing is the edge between
the front bearing and the front outer face, where the rounding we toleranced
on the shaft will interface.

348 Find your way in GPS

 Figure A–59:  Edge tolerances for the housing

We have chosen only to use an edge tolerance for an edge of undetermined

shape for this edge. It requires that between 1,2 mm and 1,5 mm of material
is missing at the edge. We could have used a radius tolerance like we did for
the shaft. In that case we should have chosen a larger radius for the tolerance
on the housing. We could also have chosen a chamfer. In that case we should
also have made sure that there would not have been interference between the
two components.
One of these latter tolerances may have been a better choice in this case, but
using the edge tolerance of undetermined shape allows me to show how it is
indicated in the general edge tolerance indication that other individual edge
tolerances can be found in the drawing. Otherwise the general edge toler-
ances for the housing are the same as those for the shaft.

Find your way in GPS 349

 A.9 The finished drawings
Figures A–60 and A–61 show the finished drawings with all requirements for
the functional features of the shaft and housing.
There are at least two things that are special about drawings toleranced ac-
cording to the IfGPS 8-point tolerancing procedure. The first is that there are
several requirements for many of the features. As discussed throughout the
example, this technique allows for very detailed requirements for each indi-
vidual feature.
By having several tolerances for each feature, you can combine small toler-
ances with many unlocked degrees of freedom with larger tolerances with few
or no unlocked degrees of freedom. This makes the component easier and
cheaper to manufacture than it would have been, had you just used the small-
est tolerance value for the fixed tolerance, see A.11.
The other special thing is that where there are several tolerances to the same
feature, they either 1) reference the same datum system, except the tertiary
or in some cases the tertiary and secondary datum are missing because they
are unnecessary, or 2) for the datum features in the local datum system, there
is one large fixed tolerance that references the global datum system and a
number of smaller tolerances to the local datum system. See for example the
bearing surfaces and the front outer face on the housing in figure A–61.

350 Find your way in GPS

Figure A–60:  Finished drawing of the shaft

Find your way in GPS 351

 Figure A–61:  Finished drawing of the housing

352 Find your way in GPS

A.10 Non-functional features
A typical critique of the IfGPS 8-point procedure in real design and toleran-
cing situations, is that it takes a lot of time and requires a lot of consideration
by the designer.
This is why we identified the functional features in figures A–6 and A–7. The
purpose of this was to allow us to concentrate on these features and only use
the procedure on these features. This is a compromise that allows us to get the
majority of the advantages of the procedure without spending more time than
absolutely necessary.
Of course, opinions can differ about which features are functional in a given
situation, so you can come to different conclusions about which features are
included in the procedure.
Companies that introduce the procedure usually start by including only a few
features on each component, but as they gain experience and become more
comfortable with the procedure, they start using it on a larger and larger per-
centage of the features.
In any case, the features that are not toleranced using the 8-point procedure
have to be toleranced in some other way.
Since these features are the non-functional features, it is often acceptable to
tolerance them with ±tolerances, even though these tolerances are not unam-
biguous, as discussed in chapter 2.
Figures A–62 and A–63 show the tolerancing of the non-functional features
of the shaft and housing, respectively.

Find your way in GPS 353

 Figure A–62:  Tolerancing of the non-functional features
of the shaft

354 Find your way in GPS

Figure A–63:  Tolerancing of the non-functional features of the
housing

Find your way in GPS 355

 A.11 The expensive way
A typical critique, mostly from suppliers who do not understand GPS, is that
the component will be more expensive to produce when there are so many
complicated tolerances in the drawing.
This critique is based on a fundamental misunderstanding of GPS. Using the
many tolerances enables you to create very specific requirements for the com-
ponent. This allows you to assign larger values, which are easier to work with,
to the tolerances that are otherwise hard to hold, because they have few or no
unlocked degrees of freedom.
Figure A–64 shows a “simplification” of the drawing in figure A–60. If the
drawing in figure A–60 ensures the function of the shaft, then the drawing
in figure A–64 will also ensure the function of the shaft. The simplification
means that there are 11 geometrical tolerances instead of 19 and the individ-
ual surface texture tolerances have been replaced by a general surface texture
tolerance.
But if you look at the drawing in detail, you will see that where several tol-
erances gave requirements for the same feature in figure A–60, it has been
necessary to take the one that has the fewest unlocked degrees of freedom
and use the smallest tolerance value for the feature. If you compare the two
drawings, you will see that some of the tolerances have been tightened dra-
matically.
We have also had to take the finest surface texture requirement and make it
the general tolerance for the entire shaft.

356 Find your way in GPS

 Figure A–64:  “Simplified” functional drawing of the shaft

Figure A–65 shows a “simplification” of the drawing in figure A–61. Similar to

the drawings for the shaft, if the drawing in figure A–61 ensures the function
of the housing, then the drawing in figure A–65 will also ensure the function
of the housing. The simplification means that there are 10 geometrical toler-
ances instead of 15 and the M indications have been removed. All individual
surface texture tolerances, except those to the bearing surfaces, have been re-
placed by a general surface texture tolerance. Finally, the local datum system
has been removed.

Find your way in GPS 357

 If you look at this drawing in detail, you will see the same restrictions as you
saw when comparing figure A–60 and A–64. In addition, by removing M we
have had to split the tolerance between the size and the geometrical tolerance,
which makes the housing harder to make.
Finally, by removing the local datum system, we have had to tighten some tol-
erances very dramatically. This is what happens, if you do not use functional
datum systems. You typically see this when you try to combine the functional
drawing and the manufacturing drawing. In this case you often end up toler-
ancing the component based on how it is made instead of how it functions.

358 Find your way in GPS

Figure A–65:  “Simplified” functional drawing of the
housing

Find your way in GPS 359

 For example, if you look at the mounting holes and the counterbores, you
would usually machine them in the same operation. From a manufacturing
point of view (and a measurement point of view) it would be tempting to tol-
erance them in the same datum system, i.e. the one defined by the fixturing
of the component during machining, instead of tolerancing the counterbores
using the mounting holes as datums. I have simulated this in figure A–65 by
tolerancing the counterbores in the global coordinate system.
If we look at the situation in figure A–61 before we combined the tolerances,
we wanted the mounting holes and the counterbores to be coaxial within
0,1 mm, i.e. their axes could be separated by 0,05 mm, because the tolerance
value is the diameter of the tolerance zone. Figure A–66 shows the tolerance,
where each mounting hole is used as datum for its own counterbore.

Figure A–66:  Tolerances with functional datums

If instead we tolerance both the mounting holes and the counterbores in the
same datum system (datum H-H is missing from the mounting hole toler-
ance indication, because the 3× indication shows that it is a TED pattern,
so the datum is implicitly there), we have to split the tolerance. Because the
mounting holes have a 0,02 mm tolerance, we have to reduce the counterbore
tolerances to 0,08 mm, because a given mounting hole can be 0,01 mm away
from its nominal position, so the corresponding counterbore can only be al-
lowed to be off from the nominal position by 0,04 mm in the opposite direc-
tion, see figure A–67, if we still have to live up to the functional requirements.

360 Find your way in GPS

 Figure A–67:  Reduction of tolerances with non-functional
datums

In general, it turns out that if you do not use the functional datum system,
you have to use two tolerances to define a relationship instead of one. In this
case we have one tolerance for the mounting hole in the non-functional da-
tum system and another for the counterbore. If we had used the functional
datum system and toleranced the counterbore directly from the mounting
hole, as in figure A–66, there would only have been one tolerance. In this
case we split a 0,1 mm tolerance into one tolerance of 0,02 mm and another
of 0,08 mm. Such a reduction in tolerance values makes the component more
expensive to manufacture.

A.12 Summary
In this appendix we have gone through an example of the use of the IfGPS
8-point tolerancing procedure for a pair of interacting components, a shaft
and a housing.
Both the components and their described functions are quite simple. This is
to better be able to focus on the systematic tolerancing approach instead of
discussing the peculiarities of a more complex component function.
The 8-point procedure is a systematic method for tolerancing. The precondi-
tion for using the procedure is that in addition to understanding GPS, you
understand the function of the component. In most cases, you will look at
all the components that are part of a product or a function simultaneously to
ensure the correct interplay between the components.

Find your way in GPS 361

 The first point in the procedure is to define the datum system(s) of the com-
ponent based on its function. The datum features are toleranced relative to
each other, so the datum systems become stable and repeatable.
When the datum systems are defined, the functional features of the compo-
nent can be laid out with TED dimensions.
The second point in the procedure is to tolerance all the features of size that
are parts of fits with size tolerances.
Next, the functional features are toleranced with fixed tolerances, related mo-
bile tolerances and unrelated form tolerances in points 3, 4 and 5 of the pro-
cedure.
The drawing is reviewed in point 6 of the procedure to find tolerances that
can be combined, when the size tolerances and the geometrical tolerances are
defined. It can either be size tolerances and geometrical tolerances that are
combined with M , and R , or it can be coaxiality tolerances and cylindric-
ity tolerances or concentricity tolerances and roundness tolerances that are
combined into total run-out or circular run-out tolerances, respectively.
Finally, surface texture tolerances and edge tolerances are in points 7 and 8 of
the procedure.
You can make it much easier to communicate the functional requirements for
the component by using this structured 8-point procedure to tolerance your
drawings using GPS.
When you master the 8-point procedure, you will, like the author with his
novel, be able to tell a good and interesting story with your tolerancing.

362 Find your way in GPS

Find your way in GPS 363
 Appendix B The degrees of freedom indicator

Degrees of freedom are a central concept in GPS and zone based tolerancing.
The difference between two tolerances with different symbols, e.g. position
and parallelism, or two identical tolerances, except from a different number
of datums being referenced, is that the two tolerance zones have different
unlocked degrees of freedom. Consequently, there is a difference in which
directions and orientations the tolerances lock the toleranced feature and in
which directions they can adjust to fit the toleranced feature and therefore
allow it to move.
A degrees of freedom indicator is used in many figures in this book, especially
in chapters 3 and 4 and in appendix A, to show which degrees of freedom are
free, locked and redundant for a given tolerance zone.
This degrees of freedom indicator is not standardised and is only used here as
a tool to communicate the degrees of freedom situation for tolerance zones.
The rules used to draw the degrees of freedom indicator, and which therefore
should be used to interpret it, are covered in this appendix.

Unlocked degrees of freedom

We discussed the 6 degrees of freedom that all components and tolerance
zones have in chapter 5. Figure B–1 shows a degrees of freedom indicator,
where all six degrees of freedom are unlocked.

Figure B–1:  Degrees of freedom indicator indicating that

all six degrees of freedom are unlocked

As you can see, the translatory (linear) degrees of freedom are represented by
red lines with an arrowhead at both ends, and the rotatory (angular) degrees
of freedom are represented by projected red circles with one arrowhead.

364 Find your way in GPS

Redundant degrees of freedom
In chapter 3 we discussed the different shapes tolerances zones can take and
which degrees of freedom are redundant for each type of tolerance zone. A
redundant degree of freedom is a degree of freedom where the tolerance zone
does not change its meaning, when it moves along this degree of freedom.
Figure B–2 shows the degrees of freedom indicator used in figure 3–5 to indi-
cate which degrees of freedom are redundant for a tolerance zone limited by
two parallel, horizontal planes.

Figure B–2:  Degrees of freedom indicator indicating the

three redundant degrees of freedom for a tolerance zone
limited by two horizontal planes

As you can see, the redundant translatory degrees of freedom are represented
by black lines with a filled circle at each end, and the redundant rotatory de-
gree of freedom is represented by a projected black circle with a filled circle.
Figure 4–46 shows a flatness tolerance, i.e. a tolerance where the tolerance
zone has the same three redundant degrees of freedom as shown in figure B–2
and where the remaining three degrees of freedom are unlocked. The result-
ing degrees of freedom indicator is shown in figure B–3.

Figure B–3:  Degrees of freedom indicator indicating the

three redundant degrees of freedom and three unlocked
degrees of freedom for a flatness tolerance for a horizontal
plane feature

Figure B–3 shows how a combination of redundant and unlocked degrees of

freedom is shown with the degrees of freedom indicator. This degrees of free-
dom indicator shows that the two horizontal translatory degrees of freedom
and the rotatory degree of freedom around the vertical axis are redundant,

Find your way in GPS 365

 i.e. the tolerance zone does not change when it moves in these directions.
Additionally, the degrees of freedom indicator shows that the tolerance zone
can adjust to fit the toleranced feature by moving vertically up and down and
rotating around the two horizontal axes. Therefore the tolerance zone does
not constrain the toleranced feature in these three degrees of freedom.

Locked degrees of freedom

If a tolerance references datums, then these will lock one or more of the de-
grees of freedom of the tolerance zone. Figure B–4 shows the degrees of free-
dom indicator from figure 4–21. It applies to a parallelism tolerance for a
horizontal plane surface in relation to a plane datum.

D D

Figure B–4:  Degrees of freedom indicator indicating the

three redundant degrees of freedom, two locked degrees
of freedom and one unlocked degree of freedom for a
parallelism tolerance for a horizontal plane feature

As we can see in figure B–4, the projected circle is removed when a rotatory
degree of freedom is locked. It is replaced by a square box that indicates the
datum that locks the degree of freedom. The square box has the same colour
as datum in the figure, where the degrees of freedom indicator is used.
This degrees of freedom indicator shows the same redundant degrees of free-
dom as figure B–3. Additionally, the degrees of freedom indicator shows that
the tolerance zone can adjust to fit the toleranced feature by moving vertically
up and down, but that the rotations around the two horizontal axes are locked
by datum D.
Figure B–5 shows the degrees of freedom indicator from figure 4–9. It applies
to a position tolerance for the median line of a hole.

366 Find your way in GPS

 P

S
M

Figure B–5:  Degrees of freedom indicator indicating the

two redundant degrees of freedom and four locked degrees
of freedom for a position tolerance for the median line of a
horizontal hole

As we can see in figure B–5, the two rotations are locked by datum S. This is
symbolised the same way as in figure B–4. In addition, we can see that the
vertical movement is locked by datum P and the non-redundant horizontal
movement is locked by datum M. In both cases the arrowheads have been
removed, the line representing the axis is shown in grey, and a square box
indicating the datum that locks the degree of freedom has been added. The
square box has the same colour as datum in the figure, where the degrees of
freedom indicator is used.
Since there are no red arrows in the degrees of freedom indicator, the toler-
ance zone is completely locked by the referenced datums and cannot move.

Sets of tolerance zones

Tolerances for line features often define a set of tolerance zones. In this case
the direction of the set is indicated by a dashed blue line with a filled circle at
each end, if the direction is linear, as in figure 4–55, or with a blue dashed pro-
jected circle with a filled circle, if the direction is circular, as in figure 4–61.
Figure B–6 shows the degrees of freedom indicator from figure 4–55. It ap-
plies to a straightness tolerance for a horizontal, plane surface.

Figure B–6:  Degrees of freedom indicator indicating a set

of tolerance zones arranged in one horizontal direction,
where each tolerance zone has one redundant degree of
freedom and two unlocked degrees of freedom

Find your way in GPS 367

 Those degrees of freedom that are locked by the direction of the set of toler-
ance zones are indicated just like other locked degrees of freedom, i.e. locked
rotatory degrees of freedom are not shown, and locked translatory degrees of
freedom are shown as grey lines. In figure B–6 there are two rotatory degrees
of freedom that are locked by the direction of the set of tolerance zones.

Diameters as degrees of freedoms

In some cases, such as roundness and cylindricity, the median diameter of
the tolerance zone is an extra degree of freedom. In these cases it is indicated
by a letter D close to the rotatory degree of freedom corresponding to the
diameter direction. Figure B–7 shows the degrees of freedom indicator from
figure 4–63, where the median diameter of each tolerance zone in the set is an
extra degree of freedom.

Figure B–7:  Degrees of freedom indicator indicating a set

of tolerance zones arranged in one horizontal direction,
where each tolerance zone has one redundant rotatory
degree of freedom, two unlocked translatory degrees of
freedom and one diameter degree of freedom

Degrees of freedom locked by the direction of the leader line

In a single case in this book, there is an example, in figure 4–29, of a toler-
ance zone whose direction is locked by the direction of the leader line for the
tolerance indicator.
In this case an arrow meeting a vertical line is shown in a square at the degree
of freedom locked by the direction of the leader line. Figure B–8 shows the
degrees of freedom indicator from figure 4–29.

368 Find your way in GPS

 Figure B–8:  Degrees of freedom indicator indicating that
a rotatory degree of freedom is locked by the direction of
the leader line for the tolerance indicator

Note that in this case the direction of the leader line is only approximately,
but not unambiguously, defined on the skin model and the real component.

Summary
The degrees of freedom indicator is used in this book to indicate which de-
grees of freedom are unlocked, locked and redundant for the tolerance zones
in the explanation drawings. They hopefully help the understanding of how
tolerance zones can move and therefore the understanding of zone based tol-
erancing and GPS in general.
The degrees of freedom indicator can be useful as a tool to analyse and under-
stand a tolerance in a drawing.
If you want to draw the degrees of freedom indicator for the tolerances in
drawings, the easiest is to follow these steps:
1. Identify the redundant degrees of freedom from the shape of the tol-
erance zone. You can use chapter 3 to help determine which degrees
of freedom are redundant for the different tolerance zone shapes.
2. If the tolerance defines a set of line tolerances (straightness, round-
ness or line profile), identify the direction of the set and which de-
grees of freedom are locked by the set (typically two rotatory degrees
of freedom).
3. If the tolerance references one or more datums, identify which
degrees of freedom are locked by which datum. You can use chapter
5 to help determine which degrees of freedom the three types of da-
tums (point, line plane) can lock. Remember the priority sequence in
a datum system.
4. If the tolerance is controlled by the direction of the leader line,
identify which rotatory degrees of freedom are locked by this direc-
tion.

Find your way in GPS 369

 5. Identify the remaining degrees of freedom as unlocked.
6. If there is a variable diameter (roundness, cylindricity, circular
run-out, total run-out), indicate in which direction the diameter is
variable.

370 Find your way in GPS

Index

A Degrees of freedom indicator, 364-370

Derived features, 13, 59
Acceptance rules, 255
Dimensional tolerancing, 35, 50, 234-
Angularity, 114-119
236
Association, 15
Dimensions other than size, 49
Average parameters, 249-250
Drawing indications, 35, 257
Duality principle, 14, 17
B
Basic GPS specification, 23 E
Basic parameters, 248
Edges, 275-290
Envelope cylinder, 201
C Envelope requirement, 42, 44, 48, 52,
Circular run-out, 157 192, 201, 208, 215
Coaxiality and concentricity, 95-96 Evaluation, 15
Collection, 15 Extraction, 15
Common datum, 179
Complementary GPS standards, 32 F
Conformance, 29
Feature principle, 22
Construction, 15
Features, 10
Correlation ambiguity, 26, 34
Functional features, 296-298
Cut-off filter, 23, 258-263 Non-functional features, 353-355
Cut-off length, 243-247 Features of size, 36, 323-324
Cylindricity, 133, 302 Filtration, 15
Fits, 235
D Flatness, 129, 307
Datums and datum systems, 163, 299- Functional control principle, 24
323
Datums in the tolerance indicator, 182 G
Datum targets, 177 General GPS standards, 31
Decimal principle, 23 General specification principle, 25
Default, 146 Global size, 38
Default principle, 23 GPS matrix system, 17
Definitive drawing principle, 20
Degrees of freedom, 59, 79, 306

Find your way in GPS 371

 H Minimum circumscribed size, 39
MMS, 196-198, 204, 206, 213
Hole distance, 56
MMVC, 196
Holes, 169, 180
Hole tolerance, 235
N
I Nominal model, 13, 16, 33
Non-ideal feature, 92
Ideal feature, 90
IfGPS 8-point procedure, 291, 299, 353
Implementation uncertainty, 16 O
Independency principle, 22 Operators, 16, 33
Inside cylinders, 169 Outside cylinders, 169
Integral features, 13, 75, 80
ISO basic GPS specification, 23 P
ISO code, 46
Parallelism, 101-106
ISO tolerance codes, 44
Partition, 15
IT grades, 45, 234
Perpendicular axes, 62
Perpendicularity, 108-111, 303, 311
L Position, 83, 120-122
Least material requirement, 215 Primary profile parameters, 248
least material size. See LMS Pt, 255
Letter case, 45 Pt. See Primary profile parameters
Line profile, 93, 125, 149
Line requirement, 91 R
LMS, 196, 208, 213
Ra, 250
LMVC, 217-223
Real model, 14, 33, 176
Locked degrees of freedom, 366-367
Reference condition principle, 24
Related tolerances, 78, 86, 149, 162
M Responsibility principle, 25
Machining allowance, 264 Rigid workpiece principle, 24
Macro geometry, 239 Rotational degrees of freedom, 100, 122,
Manufacturing method, 265 364-370
Material ratio curve, 266 Roughness and form error, 167, 240
Maximum inscribed size, 38 Roundness, 145
Maximum material requirement, 192 Rp, 253
maximum material size. See MMS Rq, 252-253
Measurand, 10 Rt, 254
Measurement uncertainty, 27, 33 Rv, 254
Method uncertainty, 33 Rz, 254
Micro geometry, 239

372 Find your way in GPS

S Surface texture requirement, 258, 261-
264, 345-347
Shafts, 169, 180
Surface texture tolerances, 239
Shaft tolerance, 235
Symmetry, 99-100
Size requirements, 44
Skin model, 14, 17, 33
Specification, 14 T
Specification ambiguity, 26, 34 TEDs, 184-188, 310-311, 321-323
specification operator, 33 Temperature, 24
Standards, 270-271 Theoretically exact dimensions, 184
ANSI/ASME Y14.5, 22 tolerance grades, 45, 234
ISO 1, 30 Tolerances, 16
ISO 129-1, 277, 284-287 Coaxiality tolerance, 201
ISO 286-1, 20, 44-45 Edge tolerances, 347-349
ISO 286-2, 20 Fixed tolerances, 81, 325-333
ISO 1101, 90, 277 Form tolerance, 63, 78, 149, 221
ISO 2768-1, 233-238 Geometrical tolerances, 338-343
ISO 2768-2, 233-238 Mobile related tolerances, 333-338
ISO 4288, 255, 260 Surface tolerances, 265
ISO 5459, 180 Tolerance zones, 55
ISO 8015, 19, 34, 48 Unrelated form tolerances, 336-338
ISO 8785, 274 Tolerance symbols, 77
ISO 10579, 24, 31, 229-231 Tolerancing, 46
ISO 13715, 277-283, 289 Total run-out, 152
ISO 14253-1, 20
Translatory degree of freedom, 82, 101
ISO 14405-1, 44, 48
Two-point size, 38, 44
ISO 14660-1, 12
ISO 17450-1, 31
ISO 17450-2, 31 U
ISO/IEC Guide 98-3, 30 Unrelated tolerance, 137
ISO/IEC Guide 99, 30 Unrelated tolerance zones, 128
ISO/TR 14638, 17, 19
ISO/TR 16570, 277, 285-286, 290
ISO/TS 17450-2, 23 V
The ISO 14253 series, 31 Verification, 10
The ISO 14660 series, 31 Verification operator, 16
The ISO 16610 series, 31, 258
Step height, 55 W
Straightness, 137
Surface imperfections, 270-273 Waviness parameters, 248
Surface lay and pattern, 263-265 Waviness profiles, 255, 262
Surface profile, 90, 122, 135 Wt. See Waviness profiles
Surface roughness specifications, 23

Find your way in GPS 373

374 Find your way in GPS
Word list

The following is a collection of the yellow notes found in this book.

Circular run-out
Circular run-out is the combined effect of several geometrical deviations for
features with cylindrical symmetry considered in one cross-section at a time,
as opposed to total run-out. You can say that the value expresses how far the
feature “runs out” in each cross-section relative to the datum. Circular run-
out tolerances are typically used to limit these deviations in rotating compo-
nents. There are two kinds of circular run-out tolerances:
Radial circular run-out limits the radius variation in one cross-section at a
time and can be interpreted as roundness relative to a nominally concentric
datum. Radial circular run-out allows radius deviation in the axial direction
and can therefore be used for e.g. conical features, as opposed to radial total
run-out.
Axial circular run-out controls wobbling in the axial direction of one concen-
tric cross-section (i.e. in one radial distance) at a time, relative to a datum line.
Axial circular run-out allows dish shape, as opposed to axial total run-out (or
perpendicularity), but ensures that the dish is centred and perpendicular to
the datum. It can therefore be used for non-plane surfaces.

Datum
Ideal derived feature that is used as the basis for tolerancing. A datum is al-
ways a plane, a straight line or a point.

Datum feature
Real feature from which the datum is derived.

Default
Rules that apply unless otherwise specified. For example, in some countries
50 km/h is the default speed limit in cities.

Degree of freedom
Parameter that describes the position of a component or the position or state
of a tolerance zone. All objects have six positional degrees of freedom. These
Find your way in GPS 375
 describe the location of the object as coordinates in a Cartesian coordinate
system and its orientation as rotations around the three axes in the coordinate
system. In addition, some tolerance zones have intrinsic degrees of freedom
that allow them to adjust to fit the toleranced feature. For example, the me-
dian diameter of a cylindricity tolerance.
The three translatory degrees of freedom describe the location of the object
as coordinates in a Cartesian coordinate system, and the rotatory degrees of
freedom describe the object’s orientation as rotations around the three axes
in the coordinate system.

Derived feature
Theoretical feature that is derived (constructed) from one or more integral
features. A derived feature can be ideal (without form error) or non-ideal
(with form error).

Feature of size
Feature that is characterised by a dimension, which is its size. Features of
size include (OD and ID) cylinders, spheres, pairs of opposite planes, cones
and wedges. Cylinders, spheres, and pairs of opposite planes have a linear
size (a size in millimetres). Cones and wedges have an angular size (a size in
degrees).

Fixed (related) tolerance

A tolerance where all non-redundant degrees of freedom are locked by refer-
ences to datums.

Form tolerance
A tolerance that only limits the form of the feature, but not its orientation or
location.

Ideal feature
Theoretical feature that does not have form error.

Integral feature
Real feature that forms part of the surface of the component.

Letter case
“Holes shall be large to be assembled”. Therefore capital (large) letters are
used to indicate the fundamental deviation.

376 Find your way in GPS

“Shafts shall be small to be assembled”. Therefore lower case (small) letters are
used to indicate the fundamental deviation.

Line requirement
Requirement that applies to a line or a set of lines individually. Straightness
and roundness are examples of line requirements. Line requirements are also
called 2D requirements (two dimensional requirements).

Location tolerance
A tolerance that limits the form, orientation and location (position) of the
feature.

Mobile (related) tolerance

A tolerance where some, but not all, non-redundant degrees of freedom are
locked with references to datums.

Nominal model
Theoretically perfect model of a component, as it is represented e.g. in a CAD
system.

Non-ideal feature
Theoretical or real feature that has form error.

Non-redundant degree of freedom

Degree of freedom for a tolerance zone where a change in the value of the
degree of freedom changes the tolerance zone.

Orientation tolerance
A tolerance that limits the form and orientation of the feature, but not its
location.

Real model (component)

The physical component, which has form and angular errors. The verification
of a component (the measurement) takes place on the real component.

Redundant degree of freedom

A degree of freedom for a tolerance zone where a change in the value of the
degree of freedom does not change the tolerance zone. For example, a cy-
lindrical tolerance zone has two redundant degrees of freedom; movement
along the axis of the tolerance zone and rotation around the axis.

Find your way in GPS 377

 Related tolerance
A tolerance that references datums.

Rotational degree of freedom

A degree of freedom that allows rotation around an axis. As opposed to a
translatory degree of freedom.

Skin model
Theoretical model of a component, which includes form and angular errors.
The specification of a component is defined on the skin model.

Surface requirement
A requirement that applies to a surface, as opposed to one or more lines. Flat-
ness and cylindricity are examples of surface requirements. Surface require-
ments are also called 3D requirements (three dimensional requirements).

Total run-out
Total run-out is the combined effect of several geometrical deviations for fea-
tures with cylindrical symmetry considered for the entire feature simultane-
ously. You can say that the value expresses how far the feature “runs out”
relative to the datum. Total run-out tolerances are typically used to limit these
deviations in rotating components. There are two kinds of total run-out toler-
ances:
Radial total run-out limits the radius variation in the entire feature and can
be interpreted as cylindricity relative to a nominally coaxial datum line. Ra-
dial total run-out can therefore only be used for cylindrical features.
Axial total run-out controls wobbling (like a loosely attached car wheel) in
the axial direction for the entire feature simultaneously and is the same as
perpendicularity relative to a datum line. Axial total run-out can therefore
only be used for plane surfaces and can always be replaced by perpendicular-
ity.

Translatory degree of freedom

A degree of freedom that allows linear motion in one direction. As opposed
to a rotatory degree of freedom.

Unrelated tolerance
A tolerance that does not references datums.

378 Find your way in GPS

Dr. Henrik S. Nielsen,
the author of this text, is the chairman of ISO technical committee,
ISO/TC 213, Dimensional and geometrical product specifications and
verification, which is responsible for developing the ISO GPS standards.
Dr. Nielsen has taught and written extensively about GPS and has
participated in GPS related standardisation work since 1987.
The ISO Geometrical Product Specifications Handbook
This book provides an introduction to Geometrical Product Specifications (GPS), the global language for
specifying requirements for the geometry of components. Through easy-to-understand colour illustrations
and specific examples, it guides the reader through the basic rules for interpreting the graphic GPS language
and provides a step-by-step procedure for tolerancing your components and products using GPS.
The book also shows how everyone involved in product realisation, from idea conception to manufacturing and
verification, can benefit from the use of GPS. This is because GPS tolerances can express functional requirements
more precisely and therefore be made larger, so that components become less expensive to produce.

Danish Standards
Kollegievej 6
DK - 2920 Charlottenlund
Denmark
Tel. +45 3996 62 12
Fax +45 3996 61 02
E-mail [email protected]
Web www.ds.dk

ISO Central Secretariat

1, chemin de la Voie-Creuse
CH - 1211 Genève 20
Switzerland
Tel. +41 22 749 01 11
Fax +41 22 733 34 30
E-mail [email protected]
Web www.iso.org

ISBN 978-87-7310-721-8

9 788773 107218
© ISO/Danish Standards – January 2012
First edition – Printed in Denmark