AGMA 915- 2- A05

AMERICAN GEAR MANUFACTURERS ASSOCIATION

Inspection Practices - Part 2: Cylindrical

Gears - Radial Measurements
AGMA 915- 2- A05

AGMA INFORMATION SHEET

(This Information Sheet is NOT an AGMA Standard)
 American Inspection Practices - Part 2: Cylindrical Gears - Radial Measurements
AGMA 915--2--A05
Gear
Manufacturers CAUTION NOTICE: AGMA technical publications are subject to constant improvement,
revision or withdrawal as dictated by experience. Any person who refers to any AGMA
Association
technical publication should be sure that the publication is the latest available from the As-
sociation on the subject matter.
[Tables or other self--supporting sections may be quoted or extracted. Credit lines should
read: Extracted from AGMA 915--2--A05, Inspection Practices -- Part 2: Cylindrical Gears
-- Radial Measurements, with the permission of the publisher, the American Gear
Manufacturers Association, 500 Montgomery Street, Suite 350, Alexandria, Virginia
22314.]
Approved May 3, 2005
ABSTRACT
This information sheet discusses inspection of cylindrical involute gears using the radial (double flank) compos-
ite method, with recommended practices detailed. Also included is a clause on runout and eccentricity
measurement methods. This information sheet is a supplement to the standard ANSI/AGMA 2015--2--AXX.
Published by

American Gear Manufacturers Association

500 Montgomery Street, Suite 350, Alexandria, Virginia 22314
Copyright © 2005 by American Gear Manufacturers Association
All rights reserved.
No part of this publication may be reproduced in any form, in an electronic
retrieval system or otherwise, without prior written permission of the publisher.
Printed in the United States of America
ISBN: 1--55589--843--2

ii
AMERICAN GEAR MANUFACTURERS ASSOCIATION AGMA 915--2--A05

Contents
Page
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv
1 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Normative references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
3 Symbols, corresponding terms and definitions . . . . . . . . . . . . . . . . . . . . . . . . . . 1
4 Measurement of radial composite deviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
5 Tooth thickness measurement with radial composite measurement . . . . . . . 12
6 Verification of master gears and fixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
7 Runout and eccentricity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Figures
1 Principle of measuring radial composite deviations . . . . . . . . . . . . . . . . . . . . . . 3
2 Radial composite deviation diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3 Interpretation of radial composite deviation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4 Strip chart of double flank composite test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
5a Double flank composite test, low number of teeth (12 tooth gear) . . . . . . . . . . 8
5b Double flank composite test, high number of teeth (30 tooth gear) . . . . . . . . . 9
6a Total composite deviation of 30 tooth gear (unfiltered) . . . . . . . . . . . . . . . . . . . 10
6b Long term component (30 tooth gear) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
6c Short term component (30 tooth gear) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
7 Manual interpretation of composite test (12 tooth gear) . . . . . . . . . . . . . . . . . . 11
8 Complex deviations with first order removed (one revolution) . . . . . . . . . . . . 12
9 Radial composite action test measurement of tooth thickness . . . . . . . . . . . . 14
10 Principle of measuring radial runout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
11 Anvil size for measuring radial runout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
12 Runout from coordinate measuring machine . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
13 Runout diagram of a gear with 16 teeth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
14 Runout and pitch deviations of an eccentric gear . . . . . . . . . . . . . . . . . . . . . . . 21
15 Gear with zero runout, but with considerable pitch and cumulative pitch
deviations (all space widths are equal) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
16 Gear with pitch and cumulative pitch deviations and zero runout . . . . . . . . . . 22
17 Actual gear with little runout and substantial cumulative pitch deviation . . . . 23
18 Runout measurement with a rider when all space widths are equal and
pitch deviations are present . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Tables
1 Symbols and terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 Recommended checking load for metallic gears with 2.5 mm face width . . . . 7

© 2005 AGMA ---- All rights reserved iii

AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

Foreword
This Information Sheet, AGMA 915--2--A05, Inspection Practices -- Part 2: Cylindrical
Gears -- Radial Measurements, is provided for informational purposes and is intended for
use with Standard ANSI/AGMA 2015--2--AXX, Accuracy Classification System -- Radial
Measurements for Cylindrical Gears.
AGMA 915--2--A05 replaces AGMA ISO 10064--2, Cylindrical Gears -- Code of Inspection
Practice -- Part 2: Inspection Related to Radial Composite Deviations, Runout, Tooth
Thickness and Backlash, and the information on similar subjects as covered in AGMA
2000--A88, Gear Classification and Inspection Handbook -- Tolerances and Measuring
Methods for Unassembled Spur and Helical Gears.
The user of this Information Sheet is alerted that differences exist between AGMA
2000--A88, AGMA ISO 10064--2 and this document. This includes that measuring methods
refer to an accuracy grade numbering system that is reversed, such that the smallest
number represents the smallest tolerance. Therefore, the user of this information sheet
must be very careful when comparing measurement methods formerly specified using
AGMA 2000--A88 or AGMA ISO 10064--2.
The first draft of AGMA 915--2--A05 was made in March, 1999. It was approved by the
Technical Division Executive Committee (TDEC) in May, 2005.
Suggestions for improvement of this document will be welcome. They should be sent to the
American Gear Manufacturers Association, 500 Montgomery Street, Suite 350, Alexandria,
Virginia 22314.

iv © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION AGMA 915--2--A05

PERSONNEL of the AGMA Gear Accuracy Committee

Chairman: Edward Lawson . . . . . . . . . . . . . . . . . . . . . . Gleason -- M&M Precision Systems

Vice Chairman: Steve Lindley . . . . . . . . . . . . . . . . . . . Falk Corporation

ACTIVE MEMBERS
J. Clatworthy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gear Metrology, Inc.
M.E. Cowan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gleason -- M&M Precision Systems
B.L. Cox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . BWXT Y--12, LLC
R. Frazer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . University of Newcastle--Upon--Tyne
T. Griffieth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . M&M Precision Systems Corporation
T. Klaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Milwaukee Gear Company
R. Layland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Precision Gage Company, Inc.
M. May . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gleason Works
R.W. Ott . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Caterpillar, Inc.
J.M. Rinaldo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atlas Copco Compressors, Inc.
Robert E. Smith . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . R. E. Smith & Company, Inc.

© 2005 AGMA ---- All rights reserved v

AMERICAN GEAR MANUFACTURERS ASSOCIATION
American Gear Manufacturers
Association --
Inspection Practices --
Part 2: Cylindrical Gears
-- Radial Measurements
1 Scope
This information sheet constitutes a code of practice
dealing with inspection relevant to radial composite
deviations and runout of cylindrical involute gears;
i.e., with measurements referred to double flank
contact.
In providing advice on gear checking methods and
the analysis of measurement results, it supplements
standard ANSI/AGMA 2015--2--AXX, where most of
the terms used are defined.
2 References
The following standards contain provisions which,
through reference in this text, constitute provisions of
this information sheet. At the time of publication, the
editions indicated were valid. All standards are
subject to revision, and parties to agreements based
on this document are encouraged to investigate the
possibility of applying the most recent editions of the
standards indicated below.
AGMA 915--1--A02, Inspection Practices -- Part 1:
Cylindrical Gears -- Tangential Measurements
AGMA 915--3--A99, Inspection Practices -- Gear
Blanks, Shaft Center Distance and Parallelism
AGMA 935--AXX, Recommendations Relative to
the Evaluation of Radial Composite Gear Double
Flank Testers
ANSI/AGMA 1012--G05, Gear Nomenclature, Defi-
nition of Terms with Symbols
© 2005 AGMA ---- All rights reserved
AGMA 915--2--A05
ANSI/AGMA 2002--B88, Tooth Thickness
Specification and Measurement
ANSI/AGMA 2015--1--A01, Accuracy Classification
System -- Tangential Measurements for Cylindrical
Gears
ANSI/AGMA 2015--2--AXX, Accuracy Classification
System -- Radial Measurements for Cylindrical
Gears
ANSI/AGMA 2116--AXX, Evaluation of Double
Flank Testers for Radial Composite Measurement
of Gears
ISO/TR 10064--2:1996, Cylindrical gears -- Code of
inspection practice -- Part 2: Inspection related to
radial composite deviations, runout, tooth thickness
and backlash
3 Symbols, corresponding terms and
definitions
3.1 Symbols and terms
The symbols and terms used throughout this infor-
mation sheet are in basic agreement with the
symbols and terms given in AGMA 900--G00, Style
Manual for the Preparation of Standards, and
ANSI/AGMA 1012--G05, Gear Nomenclature, Defi-
nition of Terms with Symbols. In all cases, the first
time that each symbol is introduced, it is defined and
discussed in detail.
NOTE: The symbols and definitions used in this infor-
mation sheet may differ from other AGMA standards.
The user should not assume that familiar symbols can
be used without a careful study of their definitions.
The symbols and terms are listed in alphabetical
order by symbol in table 1.
3.2 Definitions
The terms used, wherever applicable, conform to
ANSI/AGMA 1012--G05 and ANSI/AGMA
2015--2--AXX.
The reference axis of a component is defined by
means of datum surfaces. In most cases the axis of
the bore can be adequately represented by the axis
of the mating product arbor (see AGMA
915--3--A99).
1
AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

Table 1 -- Symbols and terms

Symbols Terms Units
a Center distance mm
ad Test center distance mm
d Diameter, reference pitch mm
db Diameter, base circle mm
Fid Radial composite deviation, total mm
Fp Total cumulative pitch deviation mm
Fr Runout mm
fe Eccentricity mm
fid Radial composite deviation, tooth--to--tooth mm
fpt Single pitch deviation mm
Lg Gage block stack height mm
mn Module, normal -- --
pb Pitch, base mm
Rr Test radius mm
s Tooth thickness mm
x Profile shift coefficient -- --
z Number of teeth -- --
α Pressure angle degrees
β Helix angle degrees
δ Prism (anvil) half angle degrees
εβ Helical overlap ratio -- --
η Tooth thickness half angle degrees
Subscripts
3 Master gear
a Arbor
b Base
n Normal plane
T Tolerance
t Transverse plane
w Product gear, operating
y Any (specified) diameter

The geometric axis of the teeth for radial compos-
ite deviation is that axis which, if used for the
measurement, would give the minimum root mean
square (rms) total radial composite deviation over a
complete revolution.
4 Measurement of radial composite
deviations
There are two composite measurement methods for
gear inspection. This information sheet contains a
description of the double flank composite action test,
which measures variations in the center distance
(radial direction) as a gear rotates in tight mesh with
a master gear. The single flank composite action test
measures transmission error in the tangential direc-
tion with gears that are not in tight mesh, and is
described in ANSI/AGMA 2015--1--A01 and AGMA
915--1--A02.
4.1 Checking principle
Radial composite deviations are checked on a
device on which pairs of gears are assembled with
one gear on a fixed spindle, the other on a spindle
carried on a slide provided with a spring arrange-
ment enabling the gears to be held radially in close
mesh (see figure 1). The variations in center

2 © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION
distance, which occur as the gears are rotated
together in tight mesh, are recorded. This recording
may be done manually while observing a dial
indicator, with a stylus on a chart, or electronically.
Gear deviations evaluated by the composite action
test are tooth--to--tooth composite deviation and total
radial composite deviation. In certain cases, func-
tional tooth thickness and radial runout can also be
evaluated.
For most inspection purposes, product gears are
tested against a master gear. Measured composite
errors always include deviations associated with the
master gear. Minimizing master gear deviations
allows more of the tolerance for errors in the product
gears.
The total radial composite deviation, Fid, of the gear
under inspection is equal to the maximum variation
Z backlash
master gear
During rotation, variation of center distance is measured
Figure 1 -- Principle of measuring radial composite deviations
fe
Fid
360°
z
0°
Figure 2 -- Radial composite deviation diagram
© 2005 AGMA ---- All rights reserved
AGMA 915--2--A05
of center distance during one revolution. It can be
determined from a recorded diagram. The tooth--to--
tooth radial composite deviation, fid, is equal to the
variation of center distance during rotation through
one pitch angle (see figure 2).
The tolerance values given in ANSI/AGMA
2015--2--AXX are valid for measurements made
using a master gear.
It is important to note that the accuracy and design of
the master gear, especially its engagement with the
product gear, can influence the test results. The
master gear should have sufficient depth of engage-
ment to be capable of contact with the entire
functional profile of the product gear, but should not
contact its non--functional or root parts. Such
contact can be avoided when the master gear teeth
are thick enough to compensate for the product gear
backlash allowance.
mesh
without
view Z (enlarged)
measuring
direction
product gear
Maximum
value of fid
360°
3
AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

When they are to be used for the quality grading of deviation between the geometrical axis of the teeth
accurate gears, the accuracy of the master gear and and the reference axis (i.e., the bore or shaft).
the measuring procedure used should be agreed
between the manufacturer and purchaser. 4.2 The utility of radial composite deviation data

The design of the master gear shall be agreed upon
between manufacturer and purchaser. The overlap
ratio, εβ test, may influence the results of radial
composite measurements of helical gears. The
effects of profile deviations which would be evident
with spur gears may be concealed because of the
multiple tooth and diagonal contact lines with helical
gears. A helical gear face width such that εβ test is
less than or equal to 0.5 with the product gear should
be used. However, the full face width of the product
gear should be explored.
A chart recording of approximate sinusoidal form
(with amplitude 2fe) over a single revolution indicates
eccentricity, fe, of the gear teeth. Reference to figure
2 shows how such a sinusoidal curve can be drawn
on the diagram. Eccentricity of a gear is the
Radial composite deviations include components
from the combined deviations of right and left flanks.
Therefore, determination of the individual deviations
of corresponding flanks is not feasible. The mea-
surement of radial composite deviations quickly
provides information on deficiencies of quality re-
lated to the production machine, the tool, or the
product gear setup. The method is chiefly used for
carrying out checking of large quantities of product
gears, as well as fine pitch gears.
Tooth--to--tooth composite deviations occurring at
each pitch increment tend to indicate profile devi-
ations (often profile slope deviations). A large
isolated tooth--to--tooth composite deviation may
indicate a large single pitch deviation or damaged
tooth (see figure 3).

1 revolution
Runout
These are fluctuations in center distance during one revolution of the product
gear. They appear in the diagram as slowly increasing and decreasing curves
corresponding to the ratio of the gears.

damaged tooth
Pitch deviations
They are revealed in the diagram as sudden and irregular deflections of the
recording pen of varying magnitude between two adjacent teeth.

Profile deviations
The slight undulations in the curve indicate deviations of the tooth form from the
theoretical involute profile. Each wave corresponds to the period of contact of
one tooth.

Pressure angle deviations (profile slope deviation)

The chart reveals them as regularly spaced and sharp--pointed vertical deflec-
tions, whereby each deflection corresponds to the period of contact of one tooth.

Figure 3 -- Interpretation of radial composite deviation

4 © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION
With appropriate calibration of the product gear
setup and checking methods, the measuring pro-
cess can also be used to determine the center
distance at which the product gear may be meshed
with minimum backlash. See AGMA 915--3--A99 for
recommendations on shaft center distance and
parallelism of axes. Furthermore, the procedure is
useful for checking gears required to operate with
minimum backlash, since the range of functional
tooth thickness can readily be derived from the radial
composite deviations.
For the determination of an accuracy grade, the
product gear should be checked against a master
gear exploring 100% of the functional flanks. See
clause 8 of ANSI/AGMA 2015--2--AXX. The toler-
ance values of total and tooth--to--tooth radial
composite deviations to determine an accuracy
grade for product gears are given in ANSI/AGMA
2015--2--AXX. It is emphasized that because of the
simultaneous contributions from both sets of tooth
flanks, such an accuracy grade cannot be directly
related to an accuracy grade determined by inspec-
tion of individual element deviations.
4.3 Double flank composite action test data
Gear rolling fixtures indicate changes in tight mesh
center distance by either a dial indicator, or recording
devices that may produce charts. Composite action
charts are amplified traces of the measured radial
displacement of composite deviation versus product
gear rotation. Figure 2 is a typical chart showing the
content of the data for tooth--to--tooth composite
deviation and total radial composite deviation.
The deviations shown in figure 2 include the effects
of the deviations which exist in both the product gear
and the master gear. When required, the results of
composite action tests should be reported in accor-
dance with 4.3.1 and 4.3.2.
4.3.1 Tooth--to--tooth composite deviation data
Tooth--to--tooth composite deviation data, during a
composite action test, is obtained as the product
gear is rotated through any angle of 360/z. This test
indicates values which include the effects of profile,
pitch, tooth thickness, and tooth alignment devi-
ations in both the product gear and in the master
gear. There is no practical way of subtracting the
deviations in the specified (master) gear from the
recorded values. The permissible values of tooth--
to--tooth composite deviation toleranced in ANSI/
AGMA 2015--2--AXX are the maximum values as
read on a dial or from a chart for any 360_/z segment.
© 2005 AGMA ---- All rights reserved
AGMA 915--2--A05
4.3.2 Total radial composite deviation data
Total radial composite deviation data, during a
composite action test, is obtained when the product
gear is rotated through one complete revolution. The
effects of total radial composite deviation in the
specified (master) gear may be compensated for by
the following:
-- determine the total radial composite deviation,
Fid, as measured on the chart or dial indicator;
-- obtain the total radial composite deviation, Fid3, of
the master gear (obtained from calibration);
-- determine the total radial composite tolerance,
FidT, allowed on drawing.
Then the following cases apply:
If Fid ≤ FidT -- Fid3, product gear is acceptable.
If Fid > FidT + Fid3, product gear is rejected.
If neither of these conditions exist, the product gear
is in question. Compensation for deviations may be
made by phasing, which can be done by indexing the
master gear with respect to the product gear,
repeating the test and analyzing the results. The
product gear is acceptable, if the highest of the
phased readings is:
Fid ≤ FidT + Fid3
4.4 Equipment requirements for composite
action testing
Figure 1 shows a schematic diagram of a gear rolling
fixture. This figure and the following discussion is
intended to show the basic kinematic and mechani-
cal requirements of the equipment necessary to
comply with this information sheet. It is not intended
to imply that this is the only acceptable construction.
Some items to be considered, which influence the
composite action test measurements, are:
-- Minimum Runout or Wobble. Provisions should
be made for the master gear to rotate with a
minimum of runout and lateral wobble. Fixed
hardened and ground studs are generally used for
mounting of master gears with hardened bores.
Precision interference ball bushings or centers,
for use with shank type master gears, should be
considered. Any clearance between the master
gear bore or hub and its mounting stud or bushing
may be reflected in the inspection results.
-- Parallelism of Axes. The fixture should be
designed for holding the product gear on a datum
axis which is parallel to the master gear axis.
5
AGMA 915--2--A05
Some fixtures provide a means of tilting the
master gear and product gear axes in relation to
each other. Such fixtures should incorporate
provisions for accurately setting the tilt angle and
re--aligning the axis to the zero position with
precision. This requirement also implies that
provision should be made for keeping a fixed
angular relationship between the axis of the
product gear and that of the master gear during
their movement toward and away from each
other.
-- Mounting. Provision should be made for holding
the product gear in the gear rolling fixture by the
same mounting surfaces as those which will be
used in the final assembly, when those surfaces
are specified on the gear drawing. Although not
essential to the conduct of the inspection, the use
of these mounting surfaces will eliminate differ-
ences which may be due to radial and lateral
runout deviations in the mounting.
-- Maintaining Prescribed Mesh. Provision should
be made for adjusting the force keeping the
product gear and master gear in tight mesh. This
force should be uniform over the entire reading
scale. Two traditional methods of doing this are:
(a) by means of a weight, or (b) by means of a
spring.
-- Changes in Center Distances. Provision should
be made for accurately indicating the changes in
the center distance that occur during the testing.
This may be done by means of a dial indicator or
a recording device. If recording is employed, it is
desirable to have a definite relationship between
the position on the chart and a circumferential
position on either the product gear or master gear.
An accurate method of calibrating the dial indica-
tor or recording equipment over its working range
is essential.
-- Other Considerations. Additional features which
contribute to the ease of operation and the
accuracy of the results are:
-- Means of quickly and accurately setting differ-
ent center distances on the fixture.
-- Means of driving the gears mechanically at low
speed in preference to turning them by hand.
This reduces the chance that small variations
will be undetected if the gear is driven too fast
and also reduces handling of the master gear.
-- Means of protecting the equipment from con-
taminants and accidental damage.
6 © 2005 AGMA ---- All rights reserved
AMERICAN GEAR MANUFACTURERS ASSOCIATION
-- Solid bases and a dust free, temperature
controlled environment are requirements for
the measurement of gears of extreme
accuracy.
4.5 Inspection equipment
In order to achieve the most accurate economical
inspection, the procedure used and the quality of the
rolling fixtures and master gears should be selected
based on the quality of the gears to be inspected.
4.5.1 Gear rolling fixture
Any inaccuracies in the fixture will reduce the
tolerance allowed for the inspected gear. The fixture
quality and the reliability of calibration must be
compatible with the product gear tolerance,see
AGMA 935--AXX.
4.5.2 Master gears
Master gears used for composite action inspection
may be one of three types of known quality:
-- A master gear designed specifically to inspect the
composite deviation of a product gear. It normally
will assure proper and complete inspection.
-- A standard master gear of known size and outside
diameter which may be used to inspect several
different product gears of the same circular pitch
or module. Caution must be taken to assure that
acceptable gears are not being rejected because
of excessive depth of contact by an oversize
outside diameter on the master gear. Similarly,
caution must be taken to avoid the possibility of
accepting gears with a short depth of functional
profile when the master gear has an undersize
outside diameter.
-- A selected mating gear of known quality, which
should be adjudged as to the degree of complete
inspection by calculation and calibration.
4.6 Method of conducting composite inspection
The following procedure should be applied when
using the radial composite deviation test:
-- A gear rolling fixture should be calibrated as
outlined in AGMA 935--AXX.
-- The master gear should be verified in accordance
with AGMA 2015--2--AXX.
-- The gear to be inspected and the master gear are
mounted on the gear rolling fixture. If mounting
surfaces are specified, these are to be used. Set
the checking load in accordance with 4.7 and
table 2.
AMERICAN GEAR MANUFACTURERS ASSOCIATION
-- The product gear is then rotated through at least
one complete revolution, in double flank contact
with the master gear.
-- The product gear is accepted or rejected on the
basis of the specified tolerances and the method
given in 4.3. The interpretation of the recorded
chart is given in figure 2 or alternatively in figures
6, 7 and 8.
4.7 Checking spring load and mass weight
The amount of applied spring load or dead weight
(mass) is important when checking gears on a gear
rolling fixture. Excessive load on fine tooth gears of
narrow face width, or gears made of soft materials, or
on journal type gears having slender shafts, will
result in incorrect readings caused by the deflection
of the gear teeth or shaft. Conversely, too light a load
on coarse gears of relatively wide face width will
result in incorrect readings, because of deviations in
the contact between the product gear and the master
gear.
4.7.1 Recommended loads
The recommended loads between product gear and
master gear are based on tooth size, and are given in
table 2.
4.7.2 Alternate loads
The loads in table 2 were determined empirically for
metallic gears, and are based on a face width of 2.5
mm. For other face widths, the load should be
changed proportionally, and should be agreed upon
by the manufacturer and purchaser of the product
gear. The loads are based on anti--friction mount-
ings for the movable head and include the force on
the indicating device.
Table 2 -- Recommended checking load for metallic gears with 2.5 mm face width
Module Load1), kg
2.5 to less than 25.0 1.0 to 1.2
1.25 to less than 2.5 0.9 to 1.1
0.80 to less than 1.25 0.8 to 1.0
0.60 to less than 0.80 0.7 to 0.8
0.50 to less than 0.60 0.6 to 0.7
0.40 to less than 0.50 0.5 to 0.6
0.30 to less than 0.40 0.3 to 0.4
0.25 to less than 0.30 0.1 to 0.2
0.20 to less than 0.25 0.1 to 0.2
NOTES:
1) For non--metallic gears use 1/2 of the listed value.
© 2005 AGMA ---- All rights reserved
AGMA 915--2--A05
4.8 Interpretation of composite data
Double flank composite data charts are made up
primarily of information related to radial runout and
variations in tooth form.
4.8.1 Traditional interpretation
Radial composite measurements are toleranced for
total composite deviation, Fid, and tooth--to--tooth
composite deviation, fid. They have been interpreted
from the charts as shown in figure 4. The total
composite variation was read as the difference
between the highest to lowest point on the chart. The
tooth--to--tooth variation was read as the greatest
change in any 360 degree/z part of the chart.
This may be acceptable for evaluation of the final
gear quality relative to the application for some
purposes. However, it does not tell the true picture
for diagnostic purposes. For example, it doesn’t help
in the case of determining noise potential. Also, if
one is trying to evaluate the manufacturing process,
it gives a distorted picture of the tooth form that the
machine and tool is producing. Ideally, one should
be able to sort out the effects of involute variations
from runout variations. These problems should be
dealt with separately in the manufacturing process.
The problem is that the tooth--to--tooth variation is
exaggerated along the part of the runout curve that
has the greatest slope. This has the effect of
distorting the amplitude of the data relating to that
particular tooth.
For the same quality of tooth form and runout, the
tooth--to--tooth variation will be greater for a gear
with a lower number of teeth than it will for higher
numbers of teeth. See figures 5a and 5b for a
comparison.
Equivalent diametral pitch Load1), ounces
1 to 9 33 to 39
10 to 19 29 to 35
20 to 29 25 to 31
30 to 39 21 to 27
40 to 49 17 to 23
50 to 59 13 to 19
60 to 79 6 to 10
80 to 99 3 to 5
100 to 120 3 to 5
7
AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

5.0

4.0

3.0

2.0 Unfiltered tooth--to--tooth

1.0
Amplitude

0.0 Total composite

--1.0

--2.0

--3.0

--4.0

--5.0
0 1 2 3 4 5 6 7 8 9 10 11 12
Tooth number

Figure 4 -- Strip chart of double flank composite test

5.0

4.0

3.0

2.0 Unfiltered tooth--to--tooth

1.0
Amplitude

0.0

--1.0

--2.0

--3.0

--4.0

--5.0
0 1 2 3 4 5 6 7 8 9 10 11 12
Tooth number

Figure 5a -- Double flank composite test, low number of teeth (12 tooth gear)

8 © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION AGMA 915--2--A05

5.0

4.0

3.0

2.0
Unfiltered tooth--to--tooth
Amplitude

1.0

0.0

--1.0

--2.0

--3.0

--4.0

--5.0
0 5 10 15 20 25 30
Tooth number

Figure 5b -- Double flank composite test, high number of teeth (30 tooth gear)

4.8.2 Relationship between tolerances
Because of this relationship between runout and the
tooth--to--tooth variation, the previous tolerances
had unrealistic values in some cases. In previously
existing standards, the tooth--to--tooth tolerance is
about 1/2 to 1/3 of the total composite tolerance.
This has come about in order to accommodate the
distortion of tooth--to--tooth data, by runout, and
especially for low numbers of teeth. There should be
a greater difference between total and tooth--to--
tooth (fidT= 0.1 to 0.2 x FidT). This is feasible when
the tooth--to--tooth variations are separated from the
runout variations.
4.8.3 New method
The separation of tooth--to--tooth from total variation
can be done by different techniques. Electronic
filters can be either analog circuits or digital in a
computer. This results in charts as shown in figure
6a, b and c. If these methods are not available in the
measuring system, a very good approximation can
be done manually.
Manual interpretation can be done by drawing in the
upper and lower envelope of the measured data.
The upper envelope is the long term component and
the vertical distance between the upper and lower
envelope is the short term component. This is shown
in figure 7.
These methods sort out the long term component of
the data from the short term component. For double
flank composite tests, the long term component
represents radial runout, Fr, and the short term
component represents the tooth form variations, fid.
4.8.4 Additional diagnostics
Most situations with long term component variations
will be in the sinusoidal form as shown in figures 6
and 7. This is caused by eccentricity. There are
cases, however, where long term variations will
show up at higher orders, such as shown in figure 8.
This can be caused by oval shapes, triangular
shapes, etc. This is common in ring gears where
heat treat distortions occur at the location of each
bolt hole in the blank. Even the short term
component can have distortions from variations in
the tooth shape.
These higher order variations can be analyzed by
the use of Fourier analysis techniques, such as a
Fast Fourier Transform (FFT) analyzer or by digital
filtering techniques. They also can be analyzed, to
some extent, by manual techniques using the upper
and lower envelope curves as drawn in figure 8.

© 2005 AGMA ---- All rights reserved 9

AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

5.0

4.0

3.0

Total composite variation

Fid -- Double flank
2.0
Amplitude

1.0

0.0

--1.0

--2.0

--3.0

--4.0

--5.0
0 5 10 15 20 25 30
Tooth number

Figure 6a -- Total composite deviation of 30 tooth gear (unfiltered)

5.0

4.0

3.0
Long term component

2.0
Fr -- Double flank
Amplitude

1.0

0.0

--1.0

--2.0

--3.0

--4.0

--5.0
0 5 10 15 20 25 30
Tooth number

Figure 6b -- Long term component (30 tooth gear)

10 © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION AGMA 915--2--A05

5.0

Short term component

4.0

fid -- Double flank

3.0

2.0
Amplitude

1.0

0.0

--1.0

--2.0

--3.0

--4.0

--5.0
0 5 10 15 20 25 30
Tooth number

Figure 6c -- Short term component (30 tooth gear)

5.0

4.0

3.0
Total composite variation
Fid -- Double flank

Long term component

2.0
Fr -- Double flank

1.0 Short term component

(fid -- Double flank)
Amplitude

0.0

--1.0

--2.0

--3.0

--4.0

--5.0
0 1 2 3 4 5 6 7 8 9 10 11 12
Tooth number

Figure 7 -- Manual interpretation of composite test (12 tooth gear)

© 2005 AGMA ---- All rights reserved 11

AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

40

30

20

10
Amplitude

--10

--20

--30

--40
0 1 2 3 4 5 6 7 8 9 10 11 12
Tooth number
Figure 8 -- Complex deviations with first order removed (one revolution)

5 Tooth thickness measurement with 5.1 Test fixture calibration

radial composite measurement
Radial composite measurement may be used to
measure functional tooth thickness. The functional
tooth thickness includes the effects of all tooth
variations, and the radial composite action test
measures every tooth of the product gear in one
operation. It is much faster than making multiple
measurements with another method. This is the best
method for measuring tooth thickness when the
tooling can be justified.
However, this method is limited to medium and
smaller gears, since testing machines capable of
more than 500 mm center distance are rarely
available. In special circumstances testing can be
accomplished in place on the cutting machine.
Special attention must be paid to the mounting
surfaces to assure that the test performed is
representative of the gear as it will be installed.
Special machines or attachments are required for
internal gears.
Test machines must be carefully calibrated,
particularly for fine pitch and high accuracy gears.
The composite action tooth thickness measurement
method utilizes a calibrated master gear and a gear
rolling device for composite action (double flank) that
has been calibrated for center distance. The
calibration of the gear rolling device is done with the
product gear and master gear holding arbors. In the
case of a journal type gear, the calibration should be
done with a precision setup arbor that is within 10%
of the product gear length over journals. These
arbors should not exceed 0.001 mm in taper, runout,
concentricity and measurement diameter. This
center type rolling fixture is set up as follows:
-- Select a master gear and obtain the actual test
radius, Rr3.
-- Establish the test radius of each of the arbors:
Product gear arbor, Rraw, and master gear arbor,
Rra3.
-- Obtain the maximum and minimum allowable test
radii from the product gear drawing (Rrw max and
Rrw min).
-- Calculate the maximum and minimum test center
distance, ad.
ad max = R r3 + R rw max (1)
ad min = R r3 + R rw min
(2)

12 © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION
where
Rrw min is the test radius, product gear, minimum
allowable;
Rrw max is the test radius, product gear, maximum
allowable;
Rr3 is the test radius, master gear.
-- Calculate the maximum and the minimum gage
stack height, Lg.
Lg max = ad max −
Rraw + R ra3 (3)
Lg min = ad min −
R raw + Rra3 (4)
where
Lg max is the gage block stack height, maximum;
Lg min is the gage block stack height, minimum;
Rra3 is the test radius, master gear arbor;
Rraw is the test radius, product gear arbor.
-- With the test arbors in place, set the maximum
stack of gage blocks, Lg max, in place between the
arbors, and with a testing pressure equal to that
used to perform the composite action test, record
the maximum test center distance shown. Repeat
the procedure using the minimum stack of blocks,
Lg min, and record the minimum test center
distance shown by the dial or on the recording
device.
NOTE: Other methods of setting the minimum and
maximum test center distances are acceptable if they
can be shown to be of similar accuracy.
-- The diameter of arbors for bored type product
gears and master gears shall be such as to
ensure that the gears will be wrung onto their test
arbors. Arbor sets having a diameter difference
of 0.002 mm for accuracy grade 7 or better are
convenient. Ball--bushing arbors with interfer-
ence fits can also be used. It is important when
composite checking gears in this accuracy range,
to remove all possible looseness between the
arbors and bores of both master gear and
inspected gear, by one of these methods, so that
additional runout is not reflected in the composite
chart due to inaccurate mounting.
-- In the case of gears of appreciable size, the
member having the least weight should be placed
on the movable centers.
-- The product gear should be rotated through a
minimum of one complete revolution.
© 2005 AGMA ---- All rights reserved
AGMA 915--2--A05
-- The gear is to be accepted or rejected for
functional tooth thickness on the basis of all
measurements being within the limits set by the
recorded maximum and minimum test center
distance.
5.2 Calculations for radial composite action test
measurement
The following method applies to external gears.
The proportions of the master gear must be checked
for proper meshing with the product gear to be sure
that contact takes place near to the tip and true
involute form diameters. There must be clearance
between the tips and roots.
Master gears may be marked with a test radius which
is the radial distance from the center line of the
master gear to the reference pitch line of a mating
standard rack that has its tooth thickness equal to its
space width. This is also the radius at which they
would mesh with a standard mating gear having a
tooth thickness, stw, at the reference diameter, dw, of:
π dw
s tw = (5)
2 zw
Special master gears are often required for spur
gears with nonstandard proportions. Helical gears
usually require special master gears.
Master gears must be made very accurately since
any deviation in the master gear is added, in the test
results, to the deviations in the product gear.
5.2.1 Maximum test radius
The maximum test radius is based on the maximum
tooth thickness. The calculation method assumes
that the errors in the master gear are too small to
affect the test results. This requires a very accurate
master gear, if precision gears are to be measured.
If two gears are in tight mesh, the sum of their tooth
thicknesses on their operating pitch circles is equal
to the circular pitch on that circle. Also, the operating
pitch diameters of the two gears must be in
proportion to the numbers of teeth. These relation-
ships, with the fundamental tooth thickness equa-
tions, yield simultaneous equations, from which the
operating transverse pressure angle can be found.
s btw + s bt3 − p bt
inv α wt3 = (6)
d bw + d b3
where
sbtw is maximum transverse base tooth thick-
ness of product gear, mm;
13
AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

sbt3 is transverse base tooth thickness of master z3 is number of teeth in master gear.
gear, mm;
The value of the maximum center distance, ad max,is
dbw is base circle diameter of product gear, mm; given by:
m n cos α n
db3 is base circle diameter of master gear, mm;
ad =  z + z 3 (8)
max 2 cos β b cos α wt3 w
αwt3 is transverse operating pressure angle in
tight mesh, degrees; The maximum test radius, Rrw max, is:
pbt is transverse base pitch. R rw max = ad max − R r3
(9)

αwt3 can also be calculated from: where

Rr3 is the master gear test radius, mm.
⎡snwm+sn3 − π⎤
inv α wt3 =⎪ z n+ z ⎪+ inv αt (7) 5.2.2 Minimum test radius
w 3
⎣ ⎦ Figure 9 illustrates a typical radial composite action
test chart. The “trace for maximum gear” represents
where a gear which has a tooth at the maximum effective
snw is normal tooth thickness of the product gear thickness, swt max. The tolerance band for radial
at the reference diameter, mm; composite action test or test center distance must
allow the full deviation of the total radial composite
sn3 is normal tooth thickness of the master gear tolerance plus the tooth thickness tolerance. Both
at the reference diameter, mm;
components vary with the product gear size and
zw is number of teeth in product gear; accuracy.
INCREASING “a”
Effect of tooth INCREASING “s”
thickness deviation
s
wtT
2 tan α wt3 TEST RADIUS
FOR swt max

TEST RADIUS
FOR swt min

TRACE FOR TRACE FOR

MINIMUM MAXIMUM
GEAR GEAR

ONE
REVOLUTION
OF PRODUCT GEAR

RUNOUT

TOTAL COMPOSITE
DEVIATION, Fid

Figure 9 -- Radial composite action test measurement of tooth thickness

14 © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION
In the following formula for ad min , the use of αwt3 for
the minimum pressure angle is an approximation. If
greater accuracy is required, recalculate using
equations 7 or 8 and ad min , iterating for a final value.
s wtT
ad = ad max − F id − 2
(10)
min tan α wt3
where
ad min is minimum center distance;
swtT is transverse tooth thickness tolerance at
operating diameter with the master.
R rw min = ad min − R r3
(11)
s wnT d w
s wtT = (12)
cos β d
5.3 Tight mesh center distance
Equations 8 and 10 can be used to calculate
maximum and minimum tight mesh center distance
to control the functional tooth thickness.
This method is recommended when the master gear
is used to inspect product gears of non--standard
tooth proportions.
5.4 Measurement of backlash at operating
center distance (test)
This is another measure of product gear functional
tooth thickness with a master gear. The test center
distance is fixed and it must be accurately deter-
mined that the axes are parallel and in the same
plane. The backlash of the test set should be
measured in at least two places, preferably four, at
equal intervals around the gear.
The product gear is accepted for tooth thickness on
the basis that the backlash at a fixed test distance is
within the designed tolerance.
6 Verification of master gears and fixtures
This clause describes a procedure for verifying
master gears or specified gears and gear rolling
fixtures for use in performing composite action tests
(double flank).
6.1 Verification of master gears or specified
gears
Prior to verification, the master or specified gear
should be inspected to assure that it meets all of the
individual tooth and gear blank tolerances for its
quality class.
© 2005 AGMA ---- All rights reserved
AGMA 915--2--A05
Master gears may be calibrated for either of two
different measurements. The method of verification
to be given is dictated by the end use of the master,
based on comprehensive inspection.
6.1.1 Suitability of master gear
The suitability of a specific master gear for inspecting
a given design of product gear should be established
by each of the following:
-- establish that the master gear will inspect the
functional profile of the product gear;
-- establish that the tips of the product gear teeth will
not interfere with the roots of the master gear
teeth, and that the product gear teeth will not
contact below the form diameter of the master
gear;
-- establish that the tolerance grade of the master
gear is equal to, or better than, the master gear
tolerances specified in ANSI/AGMA
2015--2--AXX.
6.1.2 Verification procedure
Verification can be performed in two ways as follows:
6.1.2.1 When high quality master gear available
If a master gear of two or more tolerance grades
better than that of the master being calibrated is
available, proceed as follows:
Step 1. Mark one tooth of the high quality master.
Step 2. Mark three teeth approximately 120_
apart on the master to be calibrated.
Step 3. Mesh the marked tooth of the high quality
master with one of the marked teeth of the master
being calibrated and rotate the master gear being
calibrated through one revolution. Note the total
composite variation reading, and repeat this
procedure for each of the two remaining marked
teeth.
Step 4. From the largest reading obtained in Step
3, subtract the known value of total composite
variation of the high quality master gear. This
difference is the value of the composite variation
to be assigned to the master being calibrated.
6.1.2.2 Verification procedure for two master
gears of similar quality
If two master gears of similar quality are to be
evaluated, proceed as follows:
Step 1. Mark three teeth on each master gear at
approximately 120 degree increments and
identify each by 1, 2, 3.
15
AGMA 915--2--A05
Step 2. Mesh each pair of master gears together
on a gear rolling fixture. Starting with tooth 1 of
Master A, rotate it starting with tooth 1 of Master
B, through one full revolution of each. Next, rotate
tooth 2 of Master B with tooth 1 of Master A, and
so on for nine combinations. Note the tooth
combination that produces the maximum total
composite variation.
Step 3. If only two gears are to be evaluated,
assign to each gear one--half of the maximum
value of total composite variation, as noted in
Step 2.
Step 4. If three or more gears are to be evaluated,
mesh each gear with each of the other gears.
Select the pair and meshing combination that
exhibits the lowest maximum value of total
composite variation.
Step 5. For the pair found to have the lowest
maximum total composite variation in Step 4,
assign one--half of the maximum value found to
each member. Call these the best masters.
Step 6. For each of the remaining combinations
in which either of the best master gears noted in
Step 5 were used, assign to the unmarked
member the value of maximum total composite
variation minus the amount assigned to the best
master in Step 5.
6.2 Gear--rolling fixture verification
The verification of a gear--rolling fixture consists of
establishing the accuracy with which the fixture can
hold the product gear and the master gear in relation
to each other, and the sensitivity of its indicating or
recording mechanism. The rolling fixture should be
calibrated in accordance with ANSI/AGMA
2116--AXX.
7 Runout and eccentricity
Runout is the total variation of the distance between
a datum surface and an indicated surface, measured
perpendicular to the datum surface. In order to be
meaningful, the datum surface and the indicated
surface must be specified or identified. Typical
specified runouts are axial and radial runout.
7.1 Axial runout
Axial runout (wobble) exists when the axis of rotation
and datum indicating surface are not perpendicular.
This is normally measured in a direction parallel to
the axis of rotation.
16
AMERICAN GEAR MANUFACTURERS ASSOCIATION
7.2 Forms of radial runout
Radial runout is formed by variations in the distance
perpendicular to the axis of rotation between the
indicated surface and a datum surface. Eccentricity
and out--of--roundness are components of radial
runout.
7.2.1 Eccentricity
Eccentricity is often the principal contributor to radial
runout. It is often caused by the difference in centers
used during cutting and running (or testing), by
distortions in mounting, or by a combination of both.
7.2.2 Out--of--roundness
Out--of--roundness is the irregular radial variation
from a datum surface in a given plane of rotation,
exclusive of eccentricity.
Out--of--roundness may be caused by errors in
machine tools, cutting tools, lack of rigidity in setup,
hardness variation in the gear blank, or heat treat
distortion.
7.2.3 Indicating over a pin
Runout of the gear teeth is measured by indicating a
difference in the indicated value of the position of a
pin or ball device, placed in each tooth space,
relative to an axis of rotation.
7.2.4 Ball probe test
Radial runout can be measured by indicating the
position of a ball probe (see figure 10). Other types
of probes can be used if applicable.
anvil
or
prism
ball or cylinder
1 Fr
2
N
Figure 10 -- Principle of measuring radial
runout
© 2005 AGMA ---- All rights reserved
AMERICAN GEAR MANUFACTURERS ASSOCIATION AGMA 915--2--A05

7.2.5 Composite action test
Radial runout may be measured by observing the
change in center distance during one revolution of
the product gear and a master gear on a gear rolling
fixture (see figure 1). The gears are rolled together in
tight mesh, with one member on a movable center
which is spring or weight loaded. The readings
include variations of the reference (master) gear and
the tooth--to--tooth composite variations in the prod-
uct gear being tested. These variations should be
considered when judging the acceptability of the
product gear.
The diameter of the ball shall be selected such that it
contacts the tooth at mid--tooth depth and it should
be placed at mid--facewidth. For the calculation of
ball diameter see ANSI/AGMA 2002--A88.
7.4 Anvil size for measuring runout
The anvil size is chosen so that it contacts the flanks
on each side of the space approximately at the
reference circle. The prism half angle, δyt , can be
determined by the following approximations, where
δyt , αyt , and ηyt are meshing angles to the point of
contact on the measuring circle (see figure 11).

7.2.6 Root circle or outside diameter runout test

Runout may be measured by indicating the root

δyt
circle or the outside diameter, when the finishing tool
has machined these surfaces simultaneously with
the tooth profiles. Machining variations may affect
these measurements, but readings obtained do not
include the effects of various other items described
in 7.3.

rb
r
7.3 Measuring principle ηyt

ry
αyt
Relative to the gear reference axis, the runout, Fr, of
gear teeth is the difference between the maximum
and the minimum radial positions of a suitable probe
tip: ball, anvil, cylinder or prism, which is placed
successively in each tooth space as the gear is
rotated (see figure 10).

Radial runout, Fr, measurements may include the

effects of the following:

-- eccentricity of the datum circle relative to the

δyn
datum axis;
-- out--of--roundness of the datum circle;
-- axial runout (wobble) of gear blank relative to the
datum axis of rotation;
-- tooth alignment variation;
-- profile variation;
-- pitch variation;
-- tooth thickness variation.
If a ball, cylinder, or anvil that contacts both sides of a Figure 11 -- Anvil size for measuring radial
tooth space is used, tolerances of ANSI/AGMA runout
2015--2--AXX may be applied. In some instances, it
is desirable to use a rider that contacts both sides of
a tooth. If this is done, the tolerances are not The anvil should touch the tooth flanks at mid--face
intended to apply. width on the measuring circle with diameter dy.

© 2005 AGMA ---- All rights reserved 17

AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

measurement. CNC results are affected by helix

angle at the point of probe contact.
δ yt = α yt + η yt (13)
7.5.2 Measurement with continuously rotating
d cos α t
cos α yt = (14) product gear
dy
The anvil, in contact with both flanks of a tooth space,
tan α n
tan α t = (15) moves with rotation of the gear through a preset arc
cos β
length. Radial deviations are measured either at the
dy = d + 2 mn x (16) highest point of the arc, or at some other fixed point

η yt = 180 π s yt
π z − dy   (17)
during the passage through the arc. This is a
practical method for measuring the runout of large
gears. Measurements can be made on measuring
For external gears machines or generating machines, but care must be
taken to ensure that the reference axis of the gear is
st =
mn π
cos β 2

+ 2 tan α n x  (18a) concentric with the axis of rotation of the machine,
and that the arc length is sufficient to indicate
For internal gears maximum deviation.

st =
mn π
cos β 2

− 2 tan α n x  (18b) 7.5.3 Approximation of runout from radial
composite deviation
For external gears Runout may be approximated from a radial compos-

s yt = d y  st
d
+ inv α t − inv α yt  (19a)
For internal gears
s yt = d y sd − inv α + inv α 
t
t yt
(19b)
dy
tan β y = tan β
d
tan δ yn = tan δ yt cos β y
7.5 Measuring runout
The simple nature of the measurement permits a
wide range of choices of measuring equipment and
degree of automation. Some methods are briefly
described in the following paragraphs.
7.5.1 Measurement with intermittent indexing of
the product gear
A simple method in which the gear is intermittently
rotated by hand is often used for small gears. The
probe, placed in successive tooth spaces, is brought
into line for measurement and recording of any
deviation of radial position relative to a datum radial
setting. When indexing and alignment are affected
by an indexing device, the gauging instrument must
have sufficient lateral movement to take into account
the effects on alignment of pitch and helix deviations.
This freedom of movement is necessary to ensure
contact between the gauging equipment and both
tooth flanks.
Multi--coordinate numerical control (CNC) measur-
ing machines may also be used for this method of
ite test as 2fe (see 4.2), by observing the change in
center distance during one revolution of the product
gear and a master gear on a gear rolling fixture (see
figures 1 and 2). The gears are rolled together in
tight mesh, with one member on a movable center
which is spring or weight loaded. The readings
include variations of the reference (master) gear and
(20) the deviations in the product gear being tested.
These should be considered when judging the
(21)
acceptability of the product gear.
7.5.4 Measuring with coordinate measuring
machine
When using coordinate measuring machines, runout
and pitch can be measured simultaneously. Two
methods are described.
a) Measurement with 2 flank contact. The probing
sphere with an appropriate diameter is moved
inside the tooth space until 2 flank contact is
realized. Depending on the device and the gear
parameters, the measurement can be produced
with a rotating table or without one, by means of
an axis parallel probe or a star probe. See figure
12.
If a probe with a standard diameter is used, the
runout deviation in every tooth space has to be
recalculated for the diameter given in the drawing.
Considering the same pitch deviation in the tooth
space, the recorded runout deviation depends on
the diameter used in centering the sphere.
Because of the changing profile angle at the
touching points, a smaller probe is more sensitive
than a bigger one and gives greater deviation.

18 © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION AGMA 915--2--A05

a) Runout test with rotating table (four axes) b) Runout test without rotating table (three axes)
and axis parallel probe with star probe
Figure 12 -- Runout from coordinate measuring machine

b) Measurement with 1 flank contact. A probe with 7.6 Evaluation of measurement

a small diameter is moved inside the tooth space.
The left and right flanks are probed at the
measurement circle. With these measurements
the position of a sphere with a diameter as defined
in ISO/TR 10064--2, clause 6.3 is calculated.
Depending on the device and the gear parame-
ters, the measurement can be processed with a
rotating table or without one, with an axis parallel
probe or by a star probe.
7.6.1 Runout, Fr
The runout, Fr, is, with reference to the gear axis,
equal to the algebraic difference between the
maximum and minimum values of the radial devi-
ation measured in accordance with 7.5. It is
composed of roughly twice the eccentricity, fe,
together with superimposed effects of pitch and
profile deviations of the gear (see figure 13).

20

18

16

14 Fr

12
micrometers

fe

10

6 fe

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1
Tooth space number
Figure 13 -- Runout diagram of a gear with 16 teeth

© 2005 AGMA ---- All rights reserved 19

AGMA 915--2--A05
7.6.2 Eccentricity, fe
A diagram showing runout measured is shown in
figure 13. The sinusoidal component of the curve
roughly drawn by hand, or calculated by the least
squares method, indicates (in the plane of measure-
ment) the eccentricity of the teeth to the reference
axis by an amount fe.
7.7 Value of runout measurement
Control of runout of gears which are required to
operate with minimal backlash, and of master gears
to be used in the measurement of radial composite
deviations, is of particular importance.
Measurement of runout as described is not neces-
sary when the radial composite deviations of gears
are to be measured. It is clear that details of single
flank deviations such as pitch or profile deviations,
cannot be derived from measured values of runout.
For example, two gears of very different accuracy
grades, with respect to ANSI/AGMA 2015--1--A01,
can have the same value of runout. This is because
a gear contacts its mate on either right or left flanks,
whereas runout values may be influenced by simul-
taneous measurement contact with both right and
left flanks. The deviations of both flanks can have
mutually compensated influences on runout. The
extent of information which can be derived from the
measurement of runout is largely dependent on
knowledge of the machining process and the charac-
teristics of the machines.
However, when the first batch of gears produced by a
given method is inspected in detail in order to
monitor compliance with a specified accuracy grade,
variation in further production can be detected by
measuring radial composite deviations, instead of
repeating the detailed inspection.
7.8 The relation between runout and pitch
deviations
When an otherwise perfect gear has an eccentric
bore, eccentricity, fe, as in figure 14, and it rotates
20
AMERICAN GEAR MANUFACTURERS ASSOCIATION
about the axis of the bore, the runout Fr will
approximately equal 2 fe. Eccentricity causes single
pitch deviations around the circumference of the
gear with a maximum value of fpt max of
2fe[sin(180°/z)]/cos αyt. The resulting cumulative
pitch deviation also has a sinusoidal form, with a
maximum value of Fp max of 2fe/cos αyt. As shown in
figure 14, the angle between the maximum cumula-
tive pitch deviation and the “runout” is about 90°. The
approximate value of this angle is 90° + αt on the left
flanks and 90° -- αt on the right flanks. Runout,
caused by eccentricity, results in a variation in
backlash, accelerations and decelerations due to
pitch deviations.
However, when little or no runout is measured it does
not mean that no pitch deviations are present.
Machining using single indexing can create a gear as
shown in figure 15, in which all tooth spaces are
equal, resulting in no runout, while substantial pitch
and cumulative pitch deviations are present. Figure
16 shows this condition graphically. Figure 17 shows
an example of an actual gear with little runout and
relatively large cumulative pitch deviations.
This condition occurs with double flank processes,
such as with form grinding or generating grinding
(both of which index between grinding successive
tooth spaces), when the bore of the gear is
concentric with the axis of the machine table and the
indexing mechanism generates a sinusoidal cumu-
lative pitch deviation. The source of this cumulative
pitch deviation may be eccentricity of the machine
index wheel.
To reveal this condition on the gear, a modified
runout check can be applied using a “rider” as a
probe, see figure 18. The reason why this check
detects the effect of the pitch deviations is that here
the pitch deviation results in tooth thickness devi-
ations, which a rider indicates as a radial change
when contacting both flanks.
© 2005 AGMA ---- All rights reserved
AMERICAN GEAR MANUFACTURERS ASSOCIATION AGMA 915--2--A05

High point of Runout, Fr

High point of
eccentricity Theoretical pitch

Measured pitch

Low point of
eccentricity

reference
axis High point of
pitch deviation, Fp
(approximately 90°
Eccentricity, fe from high point of Fr)
Runout, Fr
2fe Total cumulative
pitch deviation, Fp

Reference circle

Measuring circle

Figure 14 -- Runout and pitch deviations of an eccentric gear

© 2005 AGMA ---- All rights reserved 21

AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

Actual centerline
Theoretical centerline
of tooth space

FpR
All space
widths are equal
FpL

Fr = 0

Figure 15 -- Gear with zero runout, but with considerable pitch and cumulative pitch deviations (all
space widths are equal)

Theoretical
gear

Actual gear

Pitch
deviation

Cumulative
pitch
deviation

Runout

Figure 16 -- Gear with pitch and cumulative pitch deviations and zero runout

22 © 2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION AGMA 915--2--A05

Pitch number
F pL = 58 mm F pR = 55 mm F r = 15 mm
Figure 17 -- Actual gear with little runout and substantial cumulative pitch deviation

Fr(s)
rider
type
A

rider type B Fr(s)

1
2
N

Figure 18 -- Runout measurement with a rider when all space widths are equal and pitch deviations
are present

© 2005 AGMA ---- All rights reserved 23

AGMA 915--2--A05 AMERICAN GEAR MANUFACTURERS ASSOCIATION

Bibliography

The following documents are either referenced in the text of AGMA 915--2--A05, Inspection Practices-- Part 2:
Cylindrical Gear -- Radial Measurements, or indicated for additional information.

AGMA 2000--A88, Gear Classification and Inspection Handbook -- Tolerances and Measuring Methods for
Unassembled Spur and Helical Gears (Including Metric Equivalents)
ISO 53: 1998, Cylindrical gears for general and heavy engineering -- Standard basic rack tooth profile
ISO 54: 1996, Cylindrical gears for general engineering and for heavy engineering -- Modules
ISO 701: 1998, International gear notation -- Symbols for geometrical data
ISO 1122--1:1998, Glossary of gear terms -- Part 1: Definitions related to geometry
ISO/TR 10064--4:1998, Cylindrical gears -- Code of inspection practice -- Part 4: Recommendations relative to
surface texture and tooth contact pattern checking

24
 PUBLISHED BY
AMERICAN GEAR MANUFACTURERS ASSOCIATION
500 MONTGOMERY STREET, SUITE 350
ALEXANDRIA, VIRGINIA 22314