You are on page 1of 31

AGMA 915- 2- A05

AMERICAN GEAR MANUFACTURERS ASSOCIATION

AGMA 915- 2- A05

Inspection Practices - Part 2: Cylindrical


Gears - Radial Measurements

AGMA INFORMATION SHEET


(This Information Sheet is NOT an AGMA Standard)

Inspection Practices - Part 2: Cylindrical Gears - Radial Measurements


American
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 Association 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) composite 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
2
3
4
5a
5b
6a
6b
6c
7
8
9
10
11
12
13
14
15
16
17
18

Principle of measuring radial composite deviations . . . . . . . . . . . . . . . . . . . . . . 3


Radial composite deviation diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Interpretation of radial composite deviation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Strip chart of double flank composite test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Double flank composite test, low number of teeth (12 tooth gear) . . . . . . . . . . 8
Double flank composite test, high number of teeth (30 tooth gear) . . . . . . . . . 9
Total composite deviation of 30 tooth gear (unfiltered) . . . . . . . . . . . . . . . . . . . 10
Long term component (30 tooth gear) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Short term component (30 tooth gear) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Manual interpretation of composite test (12 tooth gear) . . . . . . . . . . . . . . . . . . 11
Complex deviations with first order removed (one revolution) . . . . . . . . . . . . 12
Radial composite action test measurement of tooth thickness . . . . . . . . . . . . 14
Principle of measuring radial runout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Anvil size for measuring radial runout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Runout from coordinate measuring machine . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Runout diagram of a gear with 16 teeth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Runout and pitch deviations of an eccentric gear . . . . . . . . . . . . . . . . . . . . . . . 21
Gear with zero runout, but with considerable pitch and cumulative pitch
deviations (all space widths are equal) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Gear with pitch and cumulative pitch deviations and zero runout . . . . . . . . . . 22
Actual gear with little runout and substantial cumulative pitch deviation . . . . 23
Runout measurement with a rider when all space widths are equal and
pitch deviations are present . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Tables
1
2

Symbols and terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 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 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
M.E. Cowan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B.L. Cox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
R. Frazer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
T. Griffieth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
T. Klaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
R. Layland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
M. May . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
R.W. Ott . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
J.M. Rinaldo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Robert E. Smith . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2005 AGMA ---- All rights reserved

Gear Metrology, Inc.


Gleason -- M&M Precision Systems
BWXT Y--12, LLC
University of Newcastle--Upon--Tyne
M&M Precision Systems Corporation
Milwaukee Gear Company
Precision Gage Company, Inc.
Gleason Works
Caterpillar, Inc.
Atlas Copco Compressors, Inc.
R. E. Smith & Company, Inc.

AMERICAN GEAR MANUFACTURERS ASSOCIATION

AGMA 915--2--A05

American Gear Manufacturers


Association --

ANSI/AGMA
2002--B88,
Tooth
Specification and Measurement

Inspection Practices -Part 2: Cylindrical Gears


-- Radial Measurements

ANSI/AGMA 2015--2--AXX, Accuracy Classification


System -- Radial Measurements for Cylindrical
Gears

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, Definition of Terms with Symbols

2005 AGMA ---- All rights reserved

Thickness

ANSI/AGMA 2015--1--A01, Accuracy Classification


System -- Tangential 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 information 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, Definition 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 information 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).

AGMA 915--2--A05

AMERICAN GEAR MANUFACTURERS ASSOCIATION

Table 1 -- Symbols and terms


Symbols
a
Center distance

Terms

Units
mm

ad
d

Test center distance


Diameter, reference pitch

mm
mm

db

Diameter, base circle

mm

Fid

Radial composite deviation, total

mm

Fp
Fr

Total cumulative pitch deviation


Runout

mm
mm

fe

Eccentricity

mm

fid
fpt
Lg
mn

Radial composite deviation, tooth--to--tooth


Single pitch deviation
Gage block stack height
Module, normal

mm
mm
mm
-- --

pb
Rr
s
x

Pitch, base
Test radius
Tooth thickness
Profile shift coefficient

mm
mm
mm
-- --

Number of teeth

-- --

Pressure angle

degrees

Helix angle

degrees

Prism (anvil) half angle

degrees

Helical overlap ratio

-- --

Tooth thickness half angle


Subscripts
3
Master gear
a

Arbor

b
n
T
t
w
y

Base
Normal plane
Tolerance
Transverse plane
Product gear, operating
Any (specified) diameter

The geometric axis of the teeth for radial composite 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.

degrees

(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 direction 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

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 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 arrangement enabling the gears to be held radially in close
mesh (see figure 1). The variations in center

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, functional 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

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 engagement 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
backlash
view Z (enlarged)

Z
master gear

measuring
direction
product gear

During rotation, variation of center distance is measured


Figure 1 -- Principle of measuring radial composite deviations

fe
Fid
360
z

Maximum
value of fid

360
Figure 2 -- Radial composite deviation diagram

2005 AGMA ---- All rights reserved

AGMA 915--2--A05

AMERICAN GEAR MANUFACTURERS ASSOCIATION

When they are to be used for the quality grading of


accurate gears, the accuracy of the master gear and
the measuring procedure used should be agreed
between the manufacturer and purchaser.

deviation between the geometrical axis of the teeth


and the reference axis (i.e., the bore or shaft).

The design of the master gear shall be agreed upon


between manufacturer and purchaser. The overlap
ratio, 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 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.

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 measurement of radial composite deviations quickly
provides information on deficiencies of quality related 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.

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

4.2 The utility of radial composite deviation data

Tooth--to--tooth composite deviations occurring at


each pitch increment tend to indicate profile deviations (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 deflections, whereby each deflection corresponds to the period of contact of one tooth.

Figure 3 -- Interpretation of radial composite deviation

2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION

With appropriate calibration of the product gear


setup and checking methods, the measuring process 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 tolerance 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 inspection 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 accordance 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 deviations 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 mechanical 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.

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 differences 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 indicator or recording equipment over its working range
is essential.

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.

-- Other Considerations. Additional features which


contribute to the ease of operation and the
accuracy of the results are:

4.6 Method of conducting composite inspection

-- Means of quickly and accurately setting different center distances on the fixture.

-- A gear rolling fixture should be calibrated as


outlined in AGMA 935--AXX.

-- 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.

-- The master gear should be verified in accordance


with AGMA 2015--2--AXX.

-- Means of protecting the equipment from contaminants and accidental damage.

The following procedure should be applied when


using the radial composite deviation test:

-- 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.

2005 AGMA ---- All rights reserved

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 mountings for the movable head and include the force on
the indicating device.

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 doesnt 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.

Table 2 -- Recommended checking load for metallic gears with 2.5 mm face width
Module
2.5 to less than 25.0
1.25 to less than 2.5
0.80 to less than 1.25
0.60 to less than 0.80
0.50 to less than 0.60
0.40 to less than 0.50
0.30 to less than 0.40
0.25 to less than 0.30
0.20 to less than 0.25

Load1), kg
1.0 to 1.2
0.9 to 1.1
0.8 to 1.0
0.7 to 0.8
0.6 to 0.7
0.5 to 0.6
0.3 to 0.4
0.1 to 0.2
0.1 to 0.2

Equivalent diametral pitch


1 to 9
10 to 19
20 to 29
30 to 39
40 to 49
50 to 59
60 to 79
80 to 99
100 to 120

Load1), ounces
33 to 39
29 to 35
25 to 31
21 to 27
17 to 23
13 to 19
6 to 10
3 to 5
3 to 5

NOTES:
1)
For non--metallic gears use 1/2 of the listed value.

2005 AGMA ---- All rights reserved

AGMA 915--2--A05

AMERICAN GEAR MANUFACTURERS ASSOCIATION

5.0
4.0
3.0
2.0

Unfiltered tooth--to--tooth

Amplitude

1.0
Total composite

0.0
--1.0
--2.0
--3.0
--4.0
--5.0
0

5
6
Tooth number

10

11

12

Figure 4 -- Strip chart of double flank composite test

5.0
4.0
3.0
Unfiltered tooth--to--tooth

Amplitude

2.0
1.0
0.0
--1.0
--2.0
--3.0
--4.0
--5.0
0

5
6
7
Tooth number

10

11

12

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

2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION

AGMA 915--2--A05

5.0
4.0
3.0

Amplitude

2.0

Unfiltered tooth--to--tooth

1.0
0.0
--1.0
--2.0
--3.0
--4.0
--5.0
0

10

15
Tooth number

20

25

30

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

2005 AGMA ---- All rights reserved

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.

AGMA 915--2--A05

AMERICAN GEAR MANUFACTURERS ASSOCIATION

5.0
4.0
Total composite variation
Fid -- Double flank

3.0

Amplitude

2.0
1.0
0.0
--1.0
--2.0
--3.0
--4.0
--5.0
0

10
15
Tooth number

20

25

30

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

5.0
4.0

Long term component


Fr -- Double flank

3.0

Amplitude

2.0
1.0
0.0
--1.0
--2.0
--3.0
--4.0
--5.0
0

10

15
Tooth number

20

25

30

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

10

2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION

AGMA 915--2--A05

Short term component


fid -- Double flank

5.0
4.0
3.0

Amplitude

2.0
1.0
0.0
--1.0
--2.0
--3.0
--4.0
--5.0
0

10

15
Tooth number

20

25

30

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

5.0
4.0

2.0
Short term component
(fid -- Double flank)

Amplitude

1.0
0.0
--1.0

Long term component


Fr -- Double flank

Total composite variation


Fid -- Double flank

3.0

--2.0
--3.0
--4.0
--5.0
0

5
6
Tooth number

10

11

12

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

Amplitude

10
0
--10
--20
--30
--40
0

5
6
Tooth number

10

11

12

Figure 8 -- Complex deviations with first order removed (one revolution)

5 Tooth thickness measurement with


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.

12

5.1 Test fixture calibration


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

ad

min

= R r3 + R rw

min

(1)
(2)

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;
is the test radius, master gear.

Rr3

-- Calculate the maximum and the minimum gage


stack height, Lg.
Lg

max

= ad

max

Lg

min

= ad

min

Rraw + R ra3

R raw + Rra3

(3)
(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 interference 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:
s tw =

dw
2 zw

(5)

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 relationships, with the fundamental tooth thickness equations, yield simultaneous equations, from which the
operating transverse pressure angle can be found.
inv wt3 =

s btw + s bt3 p bt
d bw + d b3

(6)

where
sbtw is maximum transverse base tooth thickness of product gear, mm;

13

AGMA 915--2--A05

AMERICAN GEAR MANUFACTURERS ASSOCIATION

sbt3 is transverse base tooth thickness of master


gear, mm;
dbw is base circle diameter of product gear, mm;
db3

The value of the maximum center distance, ad max,is


given by:

is base circle diameter of master gear, mm;

wt3 is transverse operating pressure angle in


tight mesh, degrees;

is number of teeth in master gear.

z3

ad

max

R rw

wt3 can also be calculated from:

where

snwm+sn3
inv wt3 = z n+ z
+ inv t
w
3

max

Rr3

(8)

= ad

max R r3

(9)

is the master gear test radius, mm.

5.2.2 Minimum test radius

(7)

where
snw is normal tooth thickness of the product gear
at the reference diameter, mm;
sn3

is normal tooth thickness of the master gear


at the reference diameter, mm;

zw

is number of teeth in product gear;


Effect of tooth
thickness deviation
s
wtT
2 tan wt3

m n cos n
z + z 3
2 cos b cos wt3 w

The maximum test radius, Rrw max, is:

is transverse base pitch.

pbt

Figure 9 illustrates a typical radial composite action


test chart. The trace for maximum gear represents
a gear which has a tooth at the maximum effective
thickness, swt max. The tolerance band for radial
composite action test or test center distance must
allow the full deviation of the total radial composite
tolerance plus the tooth thickness tolerance. Both
components vary with the product gear size and
accuracy.
INCREASING a
INCREASING s

TEST RADIUS
FOR swt max

TEST RADIUS
FOR swt min

TRACE FOR
MINIMUM
GEAR

TRACE FOR
MAXIMUM
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 wt3 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.
ad

min

= ad

max F id 2

s wtT
tan wt3

(10)

is minimum center distance;

swtT is transverse tooth thickness tolerance at


operating diameter with the master.
R rw

min

s wtT =

= ad

min R r3

s wnT d w
cos d

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:

where
ad min

AGMA 915--2--A05

(11)
(12)

5.3 Tight mesh center distance

-- 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;

Equations 8 and 10 can be used to calculate


maximum and minimum tight mesh center distance
to control the functional tooth thickness.

-- 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.

This method is recommended when the master gear


is used to inspect product gears of non--standard
tooth proportions.

6.1.2 Verification procedure

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 determined 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

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

AMERICAN GEAR MANUFACTURERS ASSOCIATION

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.

7.2 Forms of radial runout

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.

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.

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.

7.2.2 Out--of--roundness

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.

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.

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.

7.2.3 Indicating over a pin

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.

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

Out--of--roundness is the irregular radial variation


from a datum surface in a given plane of rotation,
exclusive of eccentricity.

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

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.

ball or cylinder

Fr

7.1 Axial runout

16

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.

Figure 10 -- Principle of measuring radial


runout

2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION

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 product gear being tested. These variations should be
considered when judging the acceptability of the
product gear.

AGMA 915--2--A05

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, yt , can be
determined by the following approximations, where
yt , yt , and yt 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
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.

yt

r
ry

yt

rb

7.3 Measuring principle

yt

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
datum axis;

yn

-- 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
tooth space is used, tolerances of ANSI/AGMA
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
intended to apply.

2005 AGMA ---- All rights reserved

Figure 11 -- Anvil size for measuring radial


runout

The anvil should touch the tooth flanks at mid--face


width on the measuring circle with diameter dy.

17

AGMA 915--2--A05

AMERICAN GEAR MANUFACTURERS ASSOCIATION

yt = yt + yt

(13)

measurement. CNC results are affected by helix


angle at the point of probe contact.

d cos t
cos yt =
dy

(14)

7.5.2 Measurement with continuously rotating


product gear

tan n
cos

(15)

dy = d + 2 mn x

(16)

tan t =

s yt
yt = 180
z dy

(17)

For external gears


st =

(18a)

(18b)

mn
+ 2 tan n x
cos 2

For internal gears


st =

mn
2 tan n x
cos 2

For external gears


s yt = d y

(19a)

sd inv + inv

(19b)

st
+ inv t inv yt
d

For internal gears


s yt = d y

dy
tan
d
tan yn = tan yt cos y
tan y =

yt

(20)
(21)

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) measuring machines may also be used for this method of

18

The anvil, in contact with both flanks of a tooth space,


moves with rotation of the gear through a preset arc
length. Radial deviations are measured either at the
highest point of the arc, or at some other fixed point
during the passage through the arc. This is a
practical method for measuring the runout of large
gears. Measurements can be made on measuring
machines or generating machines, but care must be
taken to ensure that the reference axis of the gear is
concentric with the axis of rotation of the machine,
and that the arc length is sufficient to indicate
maximum deviation.
7.5.3 Approximation of runout from radial
composite deviation
Runout may be approximated from a radial composite 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
the deviations in the product gear being tested.
These should be considered when judging the
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.

2005 AGMA ---- All rights reserved

AMERICAN GEAR MANUFACTURERS ASSOCIATION

a) Runout test with rotating table (four axes)


and axis parallel probe

AGMA 915--2--A05

b) Runout test without rotating table (three axes)


with star probe

Figure 12 -- Runout from coordinate measuring machine


b) Measurement with 1 flank contact. A probe with
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 parameters, the measurement can be processed with a
rotating table or without one, with an axis parallel
probe or by a star probe.

7.6 Evaluation of measurement


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 deviation 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

micrometers

14
12

Fr
fe

10
8
fe

6
4
2
0

7 8 9 10 11 12 13 14 15 16
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 measurement) 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 necessary 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 simultaneous 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 characteristics 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 yt. The resulting cumulative
pitch deviation also has a sinusoidal form, with a
maximum value of Fp max of 2fe/cos yt. As shown in
figure 14, the angle between the maximum cumulative pitch deviation and the runout is about 90. The
approximate value of this angle is 90 + t on the left
flanks and 90 -- t 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 cumulative 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 deviations, 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
Eccentricity, fe

Runout, Fr

2fe

(approximately 90
from high point of Fr)

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

Theoretical centerline
of tooth space

All space
widths are equal

AMERICAN GEAR MANUFACTURERS ASSOCIATION

Actual centerline

FpR
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

F pL = 58 mm

AGMA 915--2--A05

Pitch number
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

Fr(s)

rider type B
1

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