GEAR DRIVE

SYSTEM S
DESIGN AND APPLICATION
 MECHANICAL ENGINEERING
A Series o f Textbooks and Reference Books

EDITORS
L. L. FAULKNER S. B. MENKES
Department o f Mechanical Engineering Department o f Mechanical Engineering
The Ohio State University The City College o f the
Columbus, Ohio City University o f New York
New York, New York

1. Spring Designer’s Handbook, by Harold Carlson

2. Computer-Aided Graphics and Design, by Daniel L. Ryan
3. Lubrication Fundamentals, by J. George Wills
4. Solar Engineering for Domestic Buildings, by William
A. Himmelman
5. Applied Engineering Mechanics: Statics and Dynamics,
by G. Boothroyd and C. Poli
6. Centrifugal Pump Clinic, by Igor J. Karassik
7. Computer-Aided Kinetics for Machine Design, by
Daniel L. Ryan
8. Plastics Products Design Handbook, Part A: Materials
and Components, edited by Edward Miller
9. Turbomachinery: Basic Theory and Applications, by
Earl Logan, Jr.
10. Vibrations of Shells and Plates, by Werner Soedel
11. Flat and Corrugated Diaphragm Design Handbook, by Mario
Di Giovanni
12. Practical Stress Analysis in Engineering Design, by Alexander Blake
13. An Introduction to the Design and Behavior of Bolted Joints, by
John H. Bickford
14. Optimal Engineering Design: Principles and Applications,
by James N. Siddall
15. Spring Manufacturing Handbook, by Harold Carlson
16. Industrial Noise Control: Fundamentals and Applications,
by Lewis H. Bell
17. Gears and Their Vibration: A Basic Approach to Understanding
Gear Noise, bv J. Derek Smith
18. Chains for Power Transmission and Material Handling: Design
and Applications Handbook, by the American Chain Association
19. Corrosion and Corrosion Protection Handbook, edited by
Philip A. Schweitzer
20. Gear Drive Systems: Design and Application, by Peter
Lynwander

OTHER VOLUMES IN PREPARATION

GEAR DRIVE
SYSTEMS
DESIGN A N D APPLICATION

Peter Lynwander
Am erican Lohmann Corporation
Hillside, N ew Jersey

M A R C E L DEKKER N ew York
Library of Congress Cataloging in Publication Data

Lynwander, Peter, [date]

Gear drive systems.

(Mechanical engineering ; 20)

Includes index.
1. Gearing. I. Title. II. Series.
TJ184.L94 1983 621.8’33 83-5278
ISBN 0-8247-1896-8

COPYRIGHT © 1983 by MARCEL DEKKER ALL RIGHTS RESERVED

Neither this book nor any part may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, microfilming, and recording, or
by any information storage and retrieval system, without permission in writing from the
publisher.

MARCEL DEKKER
270 Madison Avenue, New York, New York 10016
Current printing (last digit):
10 9 8 7 6 5 4 3 2

PRINTED IN THE UNITED STATES OF AMERICA

PREFACE

Gear drives are critical components of mechanical systems used in such diverse
industries as turbo-machinery, process, refinery, steel, construction, mining, and
marine. In all these fields there is a continuing trend toward higher reliability
and improved technology in mechanical components. Higher reliability is desired
to reduce downtime. In many applications, the cost of one day’s lost production
due to a gearbox malfunction far exceeds the initial cost of the unit; therefore,
in critical installations there is a strong emphasis on conservative design and
quality manufacture. In addition to achieving high reliability, mechanical
systems must be increasingly efficient to conserve energy. Gear manufacturers
are constantly refining their analytical, design, and manufacturing techniques to
take advantage of new technologies and provide reliable, efficient gearboxes at
minimum cost.
The purpose of this book is to present practical gearbox design and appli­
cation information to individuals responsible for the specification and operation
of mechanical systems incorporating gear drives. Sufficient theoretical informa­
tion is included to enable the engineer interested in gear analysis and design to
understand how gear units are rated and detail gear tooth geometry is defined.
The major emphasis is on parallel shaft and planetary units using spur and helical
gearing.
In addition to basic data on gear design and manufacture, such subjects as
installation, operation, maintenance, troubleshooting, failure analysis, and
economics are covered. Material on lubrication systems, bearings, couplings,
and seals is presented in order to cover all aspects of gear system operation.
Several new trends in the gear industry, due in part to the emphasis on
energy conservation, are discussed.

Hi
iv Preface

1. As mechanical equipment such as pumps, motors, compressors, turbines,

etc. are designed for higher efficiencies, rotating speeds are increased and,
therefore, higher speed transmissions are required. High speed gearing
characteristics are featured throughout the book.
2. Also, as a result of energy consciousness, there is a tendency to package
smaller mechanical systems; therefore, there is a trend developing in the
United States toward the use of planetary gear units which are far more
compact than parallel shaft designs. Included in the book is a section on
planetary gear design and application.
3. In order to achieve the highest load carrying capability in a minimum
envelope, case hardened and ground gear tooth designs are finding wide
application. This technology is covered in the book.
4. The book attempts to take a systems approach to gearbox application. It
has become apparent that gear units, when incorporated into a system of
rotating machinery, are susceptible to a variety of problems. All charac­
teristics of the drive system from the driver to the driven equipment,
including the lubrication system and accessories, can influence gearbox
operation and must be considered in the specification, installation, opera­
tion, and maintenance of the unit.
Throughout the book, standards and practices developed by the American
Gear Manufacturers Association are referred to. Successful selection, rating, and
operation of gearboxes can be accomplished by the use of AGMA publications
and the gear designer and user should be familiar with the Standards system.
The AGMA is located at 1901 North Fort Myer Drive, Arlington, Virginia 22209.,
I would like to thank Mr. Alvin Meyer and Mr. Alan Swirnow for their
assistance and comments. I am also indebted to American Lohmann Corpora­
tion, Hillside, New Jersey for its support.
Peter Lynwander
CONTENTS

Preface in

1. Types of Gear Drives: Arrangements, Tooth Forms 1

Physical Arrangement Gear Ratio Torque Loading Efficiency
Space and Weight Limitations Physical Environment References

2. Gear Tooth Design 21

The Involute Curve Gear Tooth Definitions Gear Tooth Generation
Gear Teeth in Action Rolling and Sliding Velocities Helical Gears
Internal Gears Measurement Over Balls or Wires Engineering
Drawing Format Spline Design References

3. Gearbox Rating 93
Tooth Loads Strength Rating Bending Stress Rating
Durability Rating Compressive Stress Rating AGMA Standards
for Enclosed Drive Ratings High-Speed Gearing API Standard
Sample Rating Case Scoring Flash Temperature Index
Scoring Criterion Number Minimum Film Thickness Criterion
Shaft Rating Shaft Stresses Keyways Spline Rating References

4. Bearings and Seals 145

Bearings Seals References

5. Lubrication Systems 213

Viscosity Viscosity Index Pour Point Gear Lubricants
Rust- and Oxidation-Inhibited Oils Extreme Pressure Gear Lubricants
Synthetic Lubricants Lubricant Viscosity Selection Types of
Lubrication Systems Pumps Filtration Coolers Oil Reservoir
Breather (Vent) Piping References

v
vi Contents

6. Materials and Heat Treatments 245

Hardening Fundamentals Hardenability and the Use of Alloy Steels
Steel Quality Processing of Through-Hardened Gears Carburizing
Nitriding Forging of Gear Blanks Housing Materials Hardness
Testing Nondestructive Testing Magnetic Particle Inspection
Liquid Penetrant Inspection Surface Temper Inspection References

7. Manufacturing Methods 267

Hobbing Shaping Lapping Shaving Grinding Honing
Other Methods of Gear Tooth Machining Shot Peening Inspection
Equipment Functional Inspections References

8. Gearbox Economics 287

Gear Quality Material Costs Effect of Quantity on Costs
References

9. Planetary Gear Trains 293

Analysis of Planetary Gear Arrangements Stoekicht Design
References

10. Gearbox Installation: Mounting, Alignment, Couplings 325

Couplings and System Alignment Alignment Procedure References

11. Gear Unit Operation: Testing, Startup, Condition Monitoring 335

Testing Special Tests Initial Field Startup Condition Monitoring
References

12. Maintenance and Failure Analysis: Scheduled Maintenance

Actions 383
Storage Overhaul and Spare Parts Troubleshooting References

Appendix 403

Index 407
GEAR DRIVE
SYSTEM S
DESIGN AND APPLICATION
1
TYPES OF GEAR DRIVES:
ARRANGEMENTS, TOOTH FORMS

The function of a gearbox is to transmit rotational motion from a driving prime

mover to a driven machine. The driving and driven equipment may operate at
different speeds, requiring a speed-increasing or speed-decreasing unit. The gear­
box therefore allows both machines to operate at their most efficient speeds.
Gearboxes are also used to change the sense of rotation or bridge an angle
between driving and driven machinery.
The gearbox configuration chosen for a given application is most strongly
influenced by three parameters:
Physical arrangement of the machinery
Ratio required between input and output speeds
Torque loading (combination of horsepower and speed)
Other factors that must be considered when specifying a gear drive are:
Efficiency
Space and weight limitations
Physical environment

PHYSICAL ARRANGEMENT

The location of the driving and driven equipment in the mechanical system
defines the input and output shaft geometrical relationship. Shaft arrangements
can be parallel offset, concentric, right angle, or skewed as shown in Figure 1.1.
The material presented in this book focuses on parallel offset and concentric
designs.

1
2 Gear Drive Systems

PARALLEL O FFSET CONCENTRIC

RIGHT ANGLE SKEW ED

Figure 1.1 Gearbox shaft arrangements.

In the majority of parallel offset units in use, the input and output shafts
are horizontally offset; however, vertical offsets are used and any orientation of
input to output shaft is possible. Figure 1.2 illustrates a typical horizontally
offset parallel shaft gearbox and Figure 1.3 presents a cutaway view of such a
unit. In this case there is one input shaft and one output shaft located on oppo­
site sides of the unit. There are many different options available as far as the
input and output shaft extensions are concerned. Figure 1.4 shows the various
possibilities and presents a system for defining the extensions desired on a gear­
box. There may be two inputs driving a single output, such as dual turbines
powering a large generator, or two outputs with a single input, such as an electric
motor driving a two-stage compressor. Often shaft extensions are used to drive
accessories such as pumps or starters.
The minimum amount of offset required is determined by gear tooth stress
considerations. The offset of a gearbox incorporating a single mesh, as shown in
Figure 1.3, is the sum of the pitch radii of the pinion and gear, otherwise known
as the center distance. The pitch radii must be sufficiently large to transmit the
system load. An offset greater than the minimum may be required to provide
enough space for the machinery incorporated in the system. Figure 1.5 illus­
trates an accessory drive where the input and output shafts are offset through
two meshes to separate the machinery located at these shafts.
Types o f Gear Drives 3

Figure 1.2 Parallel offset gearbox. (Courtesy of American Lohmann Corpora­

tion, Hillside, N.J.)

Figure 1.6 illustrates a gearbox with concentric input and output shafts.
The driving and driven machinery will therefore be in line. Planetary gearing
(described in the next section) has concentric shafts and is used to achieve high
ratios in minimum space. It is also possible to package parallel shaft gearing
such that the input and output shafts are in line when such a configuration is
desired. Figure 1.7 presents external and internal views of a right-angle gear
drive.
The gearboxes in Figures 1.3, 1.6, and 1.7 are foot mounted; that is, they
are meant to be bolted to a horizontal base through a flange at the bottom of
4 Gear Drive Systems

Figure 13 Parallel offset gearbox sectional view. (Courtesy of American

Lohmann Corporation, Hillside, N.J.)

the gear casing. Although this is the most common design, gearboxes can be
mounted in many other configurations and operate in attitudes other than
horizontal. Figure 1.8 illustrates a flange-mounted unit. Such a gearbox can be
operated horizontally or be vertically mounted on a horizontal base. Vertically
operating gearbox designs must have special lubrication provisions to provide
lubricant to the upper components in the unit and seal the lower end from oil
leakage. Figure 1.9 shows yet another mounting configuration. This unit is shaft
mounted with a support arm that is fixed to ground to react the gearbox housing
torque.
Types o f Gear Drives

PLAN VIEWS

f t 3 1El l- l

r r
Z3 a l= l d
R-R R-LR L-LR
[ ]=> t 3
t t

C^R^R^:]
f t t
Figure 1.4 Definition of shaft extensions. Parallel shaft-helical and herringbone
gear reducers; single, double, and triple reduction. Code; L = left; R = right;
arrows indicate line of sight to determine direction of shaft extensions; letters
preceding the hyphen refer to number and direction of highspeed shaft exten­
sions; letters following the hyphen refer to number and direction of low speed
shaft extensions. (From Ref. 1.)

Figure 1.5 Parallel offset accessory drive gearbox. (Courtesy of American

Lohmann Corporation, Hillside, N.J.)
 Gear Drive Systems

Figure 1.6 Gearbox with concentric input and output shafts. (Courtesy of
American Lohmann Corporation, Hillside, N.J.)

GEAR RATIO

There is no limit to the reduction or speed increasing ratio that can be achieved
using gearing; however, for high ratios the arrangement of the components can be
quite complex. In a simple gear mesh a maximum ratio in the order of 8:1 to
10:1 can be achieved. The amount of speed reduction or increase is simply the
ratio of the pitch diameter of the larger gear to the smaller gear. The number of
teeth in a gear pair is related to the pitch diameters, so the speed ratio can also
be calculated by dividing the larger number of teeth by the smaller. The smaller
gear is often called the pinion. To attain a ratio of 10:1, therefore, the gear must
be 10 times larger than the pinion and there usually are stress or geometrical
limitations on the pinion when this ratio is exceeded. To achieve higher ratios
with parallel shaft gearing, stages of meshes are combined as shown in Figure 1.10.
This unit has three stages of reduction and achieves ratios on the order of
100:1.
An efficient method of achieving high reduction ratios in minimum space
is the use of planetary gearing. This design, completely described in Chapter 9, is
illustrated in Figure 1.11. The high-speed sun gear meshes with a number of
Figure 1.7 Right-angle gear drive. (Courtesy of American Lohmann Corpora­
tion, Hillside, N.J.)
 Gear Drive Systems

Figure 1.8 Flange-mounted gear unit. (Courtesy of American Lohmann Cor­

poration, Hillside, N.J.)

Figure 1.9 Shaft-mounted gearbox. (Courtesy of American Lohmann Corpora­

tion, Hillside, N.J.)
Types o f Gear Drives 9

Figure 1.10 Multi-stage parallel shaft gearbox. (Courtesy of American Lohmann

Corporation, Hillside, N.J.)

planets, usually three, which in turn mesh with a ring gear. The ring gear has
internal teeth. Either the ring gear or the planet carrier rotate at the low speed
of the gear set. Occasionally, all three members are connected to rotating equip­
ment. When the low-speed shaft is either the ring gear or the planet carrier, the
ratios in a planetary gearset are:
10 Gear Drive Systems

Figure 1.11 Basic planetary gear configurations.

„ ring gear pitch diameter . ^ .

Ratio = — — ------t—t—t -------1— for rotating ring gear
sun gear pitch diameter
„ ,. 1 ring gear pitch diameter r ^ .
Ratio = 1 + — — ---- -------------- -— for rotating planet carrier
sun gear pitch diameter
Because of the multiple load path of planetary gearing the horsepower trans­
mitted is divided between several planet meshes and the gear size can be reduced
significantly compared to parallel shaft designs. Planetary stages can be linked
together to achieve high ratios, as shown in Figure 1.12. This is a three-stage
planetary gear with a ratio of 630:1. The first-stage planet carrier drives the
second-stage sun gear and the second-stage carrier drives the third-stage sun gear.
Types o f Gear Drives 11

Figure 1.12 Multistage planetary gearbox. High-speed shaft, 1; first planetary

stage, 2; second stage, 3; third stage, 4; low-speed shaft, 5; gear housing, 6.

In theory, any parallel shaft or planetary gearbox can be used either as a

speed reducer or increaser. There may be details within a gearbox, however, that
require modification if such a changeover is made. The same holds true if it is
desired to reverse the direction of rotation for which the gearbox was initially
designed. For instance, one side of the teeth may have been favored in the finish­
ing process when the initial design was manufactured. If the gearbox is used in
such a manner that the initially unloaded face is now loaded, poor tooth per­
formance may result. The direction of rotation of the input shaft with respect to
the output shaft depends on the gear design chosen. For a simple parallel shaft
gear mesh the sense of rotation will change through the mesh. A planetary
arrangement with a stationary ring gear will not change the sense of rotation
between input and output, while a rotating ring gear will turn in the opposite
sense compared to the sun gear’s rotation.

TORQUE LOADING

The size of gearbox required for a given application is dependent primarily on

how large the gear pitch diameters and face widths are. These dimensions are
determined on the basis of tooth stresses which are imposed by the transmitted
tooth load. The tooth load is simply the torque on a given gear divided by the
gear pitch radius:
torque (in .-lb)
Tooth load (lb) =
pitch radius (in.)
12 Gear Drive Systems

Torque is calculated from the horsepower transmitted and the speed of the
rotating component in question:
63.025 (hp)
Input torque (in.-lb) =----------------
input rpm
63.025 (hp)
Output torque (in.-lb) = ---------------
output rpm
When designing a gearset one cannot consider torque alone. The operating speed
of the gears has a significant effect on the design definition. As an illustration of
this point, consider a high-speed unit transmitting 2000 hp at 20,000 rpm input.
The input torque would be the same as a low-speed unit operating at 2000 rpm
input with a transmitted horsepower of 200. On a simple stress basis, if the ratio
of both gearboxes were the same, the same gearbox could be used for both appli­
cations; however, the high-speed design must differ from the low-speed design in
the following respects:
At high speeds, component geometry discrepancies such as tooth spacing error,
shafting unbalance, and so on, generate significant dynamic loading, and
these dynamic effects must be taken into account in the design process.
Also, the components experience high numbers of load cycles and are
more prone than low-speed units to fatigue failures. For all of these
reasons, high-speed components must be of high accuracy to minimize
dynamic problems.
Heat generation within the unit is proportional to speed; therefore, high-speed
units usually require pressure jet lubrication systems and external cooling
systems. Low-speed units often operate with integral splash lubrication,
the heat being dissipated through the gear casing.
The bearing design is strongly dependent on shaft speeds. Low-speed units
generally incorporate antifriction bearings, while high-speed industrial
gearboxes typically use journal bearing designs.

There is no clear demarcation between low-speed and high-speed gearing. Units

with several gear meshes may have some of each. An arbitrary definition some­
times used is that units with pinion speeds exceeding 3600 rpm or pitch line
velocities exceeding 5000 fpm are considered high speed [1].
Pitch line velocity is a measure of the peripheral speed of a gear:
7r (pitch diameter, in.) (rpm)
Pitch line velocity (fpm) = ------------- —-------------------

The pitch line velocity of a gear is a better index of speed than is rotational
velocity, since a large gear operating at a relatively low rpm may experience the
same velocity effects as a small gear operating at high rpm. Standard high-speed
gear units operate at pitch line velocities up to approximately 20,000 fpm.
Types o f Gear Drives 13

Applications exceeding this speed must be considered special and exceptional

care must be taken in their design and manufacture. Pitch line velocities of
40,000 fpm have been attained in practice.
Parallel offset or concentric shaft gearboxes incorporate gears with spur,
single helical, or double helical tooth forms. The face of a spur gear is parallel to
the axis of rotation, whereas a helical gear tooth face is at an angle, as shown in
Figure 1.13. The figure illustrates that helical gears have an overlap in the axial
direction, which results in the following advantages:
Helical gears have more face width in contact than do spur gears of the same
size; therefore, they have greater load-carrying capability.
With conventional spur gearing the load is transmitted by either one or two teeth
at any instant; thus the elastic flexibility is continuously changing as load
is transferred from single-tooth to double-tooth contact and back. With
helical gearing the load is shared between sufficient teeth to allow a
smoother transference and a more constant elastic flexibility; therefore,
helical gearing generates less noise and vibration than spur gearing.
The disadvantage of helical gearing in relation to spur gearing is that axial
thrust is generated in a helical gear, which necessitates the incorporation of a
thrust bearing on each helical gear shaft.
To take advantage of the helical gearing benefits described above, yet not
generate axial thrust loads, double helical gearing is used (Figure 1.14). The two
halves generate opposite thrust loads, which cancel out. When the two helices are
cut adjacent to one another with no gap between, the gearing is termed herring­
bone. Because helical gear thrust is proportional to the tangent of the helix
angle, single helical gears tend to have lower helix angles than do double helical
designs, where the thrust loads cancel. Typical single helical helix angles are 6 to
15°. Double helical gearsets have helix angles of up to 35°.
Another advantage of double helical gears is that the ratio of face width to
pitch diameter in each half can be held to reasonable limits. When the face
widths become longer than the pitch diameters in spur or single helical gearing it
is difficult to achieve complete tooth contact since thermal distortion, load
deflections, and manufacturing errors tend to load the gear teeth unevenly. A
double helical gear with a face width/pitch diameter ratio of 1 will have twice
the face width of a spur or single helical gear with the same L/D ratio and there­
fore greater load-carrying capability.
Double helical gearing has two disadvantages. Because the two halves of
each gear cannot be perfectly matched, one member of the gearset must be free
to float axially. This gear will be continually shifting to achieve axial force
equilibrium since the thrust loads of each half will rarely cancel exactly. This
shifting can lead to detrimental axial vibrations if tooth geometry errors are
excessive. Another potential problem with double helical gearing is that external
14 Gear Drive Systems
SPUR GEAR HELICAL GEAR

Double Tooth Contact

Oblique Line
of Contact

Advantages of Helical Gears

1. G reater load capacity

2. Sm oother operation
3. Less sensitivity to tooth errors

Figure 1.13 Comparison of spur and helical gear teeth.

thrust loads will tend to overload one helix. For instance, if a double helical
gear is attached to a gear tooth type of coupling and the coupling locks up
axially due to tooth friction, axial loads transmitted through the coupling will
be reacted by the teeth of one-half of the gearset. With single helical gears
an external axial load will either add to or subtract from the gear tooth load and
be reacted by the thrust bearing.
Gear metallurgy, although not mentioned heretofore, is one of the most
significant factors in determining gearbox size, since the strength of a gear tooth
is proportional to the hardness of the steel. Most gears are in the hardness ranges
of approximately Rc 30 to 38 or Rc 55 to 64. The region from Rc 30 to 38 is
usually termed “through-hardened,” while the range Rc 55 to 64 is almost
always “surface-hardened,” where the tooth has a hard surface case and a softer
inner core. Through-hardened gears are cut by such processes as hobbing,
Types o f Gear Drives 15

Figure 1.14 Double helical gearing.

shaping, and shaving. Surface-hardened gears are cut and then hardened. They
may be used in this state, but the more accurate surface-hardened gears are
ground after heat treatment. Spur, single helical, and double helical gearing may
be produced by any of the methods noted above. Generally, double helical
gearing is through-hardened and cut. It is possible to harden and grind double
helical gearing; however, to grind a one-piece double helical gear a large central
gap is required between the two helices to allow runout of the grinding wheel.
Gears can be ground in halves and then assembled, but this presents serious
alignment and attachment problems.
To achieve minimum envelope and maximum reliability, the latest tech­
nology utilizes single helical, hardened, and precision ground gearing. With single
helical gears the thrust load axially locates the gear shaft against the thrust bear­
ing. Bearing design has progressed to the point where thrust loads are routinely
handled either by hydrodynamic tapered land or tilting pad configurations or
an antifriction thrust bearings. Because case-hardened gears have maximum
16 Gear Drive Systems

Figure 1.15 High-speed single helical hardened and ground gearset. (Courtesy of
American Lohmann Corporation, Hillside, N.J.)

load-carrying capacity, gear size can be minimized; therefore, the ratio of face
width to diameter of a single helical gear can be held to reasonable limits. Pitch
line velocities are minimized, reducing dynamic effects. Also, the bearing span
with single helical gears is short, resulting in lower elastic deflection. Figure 1.15
illustrates a generator drive gearbox with two stages of single helical gearing. This
unit transmits 4500 hp at an input speed of 14,500 rpm. The high-speed mesh
pitch line velocity is 18,000 fpm. Gearbox weight is 3500 lb.
A single helical hardened and ground gearset can reduce by up to one-half
the envelope and weight of a through-hardened double helical gearbox with
equivalent capacity. The inherent precision of the grinding process results in
accurate tooth geometry, leading to minimum noise and vibration.

EFFICIENCY

Gearbox efficiency is a much discussed subject, but accurate values are very
difficult to determine. Analytical estimates must be confirmed by testing to gain
Types o f Gear Drives 17

a degree of confidence in the procedure. With good design and manufacturing

practice, efficiencies of 99% per mesh and better are possible. Often, lubrication
system development is required to attain the highest efficiency potential.
Power losses in a gearbox are divided between friction losses at the gear
and bearing contacts and windage losses as the rotating components churn the
oil and air. In high-speed units the churning losses may exceed the friction losses;
therefore, the type and amount of lubricant, and its introduction and evacua­
tion, are critical in terms of efficiency. Journal bearings require significantly
more oil flow than do antifriction bearings and generate higher power losses.

Figure 1.16 Compound planetary gearset. (Courtesy of American Lohmann

Corporation, Hillside, N.J.)
18 Gear Drive Systems

A reasonable estimate of efficiency for industrial gear boxes is 1 to 2%

power loss per mesh. A three-stage unit, therefore, might be expected to have
an efficiency in the range 94 to 97%. The efficiency is quoted at the design load
and speed conditions. At full speeds and lower loads the efficiency will drop off
because the churning losses will remain constant.

SPACE AND WEIGHT LIMITATIONS

There are industrial applications where gearbox space and weight is limited. For
instance, generator-drive gearboxes on offshore oil platforms or units used on
mobile equipment must have minimum envelope. To achieve small gear units,
several techniques can be used:
To minimize gear size, the highest-quality steel, case carburized and precision
ground, is incorporated in the unit.
Planetary configurations are used to achieve high ratios in small envelopes.
Figure 1.16 illustrates a compound planetary gearset which demonstrates
a very efficient use of space, in the approximate range 9:1 to 12:1.
Lightweight design techniques such as thin-wall casings and hollow shafts are
employed.
Lightweight materials such as aluminum housings are used.
Maximum application of these techniques can be found in the aerospace in­
dustry. An aircraft gearbox might handle the same design conditions as a con­
ventional unit but at one-fiftieth the weight [2].

PHYSICAL ENVIRONMENT

When specifying a gear drive, the physical environment must be addressed in the
design stage. Listed below are detrimental environments which can have an
adverse effect on lubricant, bearings, gears, or seals:
Dusty atmosphere
High ambient temperature
Wide temperature variation
High humidity
Chemical-laden atmosphere
Such environments require special consideration in the design of the gearbox
lubrication system and seals.
Types o f Gear Drives 19

REFERENCES

1. AGMA Standard 420.04, Practice for Enclosed Speed Reducers or Increasers

Using Spur, Helical, Herringbone and Spiral Bevel Gears, American Gear
Manufacturers Association, Arlington, Va., December 1975.
2. Dudley, D. W., Gear Handbook, McGraw-Hill, New York, 1962, pp. 3-5.
2
GEAR TOOTH DESIGN

The purpose of gearing is to transmit power and/or motion from one shaft to
another at a constant angular velocity. The tooth form almost universally used
is the involute, which has properties that make it particularly desirable for these
functions. It will be shown that in order to attain constant angular velocity, the
meshing tooth forms must have specific geometrical characteristics which are
easily obtained with an involute system.
In order to understand gear tooth drives it is useful to observe the dy­
namics of simpler power transmission systems, such as friction disks or belt
drives (Figure 2.1), both of which are capable of transmitting power at a
constant velocity ratio. The velocity ratio is inversely proportional to the ratio
of the diameters:

WB Da

where
W = angular velocity, rad/sec
D = diameter, in.
If Da = ^Dg it can be seen that friction disk A has to make two revolutions
for each revolution of disk B if the circumferences of the disks are rolling on one
another without slipping. Another way of looking at it is that at the point of
contact both disks have the same tangential velocity VT in inches per second,
and

21
22 Gear Drive Systems

DISK DRIVE BELT DRIVE

Figure 2.1 Friction disk and belt drives.

V - WA D A _ WB D B
V nr - ---------------- ~ -------------
1 2 2
Therefore,

Wa _ P b
WB da

Similarly, the ratio of angular velocity of sheaves A and B in the belt drive are
proportional to the ratio of diameter B to diameter A.
Disks and belt drives are power and speed limited and sometimes slip. A
more positive method of transmitting power is through gearteeth, which can be
illustrated as two cam profiles acting on one another (Figure 2.2). Theforce of
the driving cam on the driven at any instant of time acts normal to the point of
tangency of the curved surfaces. This normal line, shown as AA in Figure 2.2, is
known as the line of action. Line AA intersects a line drawn between the two
centers of rotation at point X. RA and RB are then the instantaneous pitch radii
of the two cams. The angular velocity ratio of the cams at a given instant is
inversely proportional to the ratio of the instantaneous pitch radii. For the
angular velocity ratio to remain constant, the respective pitch radii must be the
same at all points of contact. If this condition is met, the two profiles are said to
be conjugate. Two cam profiles chosen at random will rarely be conjugate;
however, given one profile a conjugate mating profile can be developed mathe­
matically. The problem is that these two conjugate profiles may not be practical
from an operating or manufacturing point of view.
This leads to one major reason why the involute curve is widely used for
gear teeth. Two mating involutes will always be conjugate and the tooth forms
relatively easy to manufacture with standardized tooling.
Gear Tooth Design 23

THE INVOLUTE CURVE

Figure 2.3 illustrates the involute curve, which may be visualized as the locus of
points generated by the end of a string which is held in tension as it is unwound
from a drum. The drum is known as the base circle and once the base circle
diameter is known, the involute curve is completely defined.Mathematically, the
involute is expressed as a vectorial angle 6 in radians:
0 = tan (j) - 0 = Inv 0
where 0 is the pressure angle at any diameter, in radians. Inorder to plot the
involute curve for a base circle of radius RB, simply assume values for the
pressure angle 0. All the terms shown in Figure 2.3 can then be calculated.
R is the radius to any point on the involute and is related to RB by the
cosine of the pressure angle:
24 Gear Drive Systems
in v o l u t e c u r v e

r B
cos 6 = —
R
Rc is the radius of curvature of any point on the involute at radius R. Inspec­
tion of Figure 2.3 reveals that R^ is also the length of string unrolled from the
base circle as the base circle rolls through an angle E; therefore,

e “Rb= Rc

Rc = \ / R2 - R |

e = E - 0 = — - tan - 1 —
RB rb
Gear Tooth Design 25
26 Gear Drive Systems

6 = tan 0 - 0
Thus, knowing the base circle radius RB and assuming values for 0, R, and 6, the
polar coordinates of the involute can be plotted. The involute in terms of the
pressure angle is useful in many gear tooth calculations, and a table of involutes
is presented in the Appendix.
Let us now mesh two involute curves together at a center distance C, as
shown in Figure 2.4. The angle 0 is now the operating pressure angle of the gear
mesh and RP and RG are the operating pitch radii. The subscripts P and G
stand for pinion and gear, with the pinion always being the smaller of the two
meshing gears. AA, the common tangent between the two base circles, is the line
of action and the two involutes are shown meshing at the pitch point. It is
important to understand that if the center distance C is increased or decreased,
the involutes will contact at different points and have different operating
pressure angles and pitch diameters. The velocity ratio, however, will not change
since it is dependent only on the ratio of the two base diameters. The relation­
ship of the base radius, pitch radius, and pressure angle is
Rbp Rb g
COS 0 = = -----
Rp Rq

The center distance C in terms of the other parameters is

r bp + r bg
C = RP + Rg =
COS 0
The insensitivity of the involute to center distance variation is another reason for
using this curve for gear teeth. Also, it can be seen that a whole system of
involute gearing can be established where within certain limits any two gears will
mesh with one another and transmit uniform rotary motion.

GEAR TOOTH DEFINITIONS

Figure 2.5 depicts a practical gear tooth. As shown, a portion of the involute
curve bounded by the outside diameter and root diameter is used as the tooth
profile. In a properly designed gear mesh the involute curve merges with the root
fillet at a point below the final contact of the mating gear. This intersection of
involute and root fillet, called the form diameter, is discussed later in the
chapter. From Figure 2.5 we can define an important parameter, the diametral
pitch, which is a measure of tooth size. The diametral pitch is the number of
teeth per inch of pitch diameter:
Gear Tooth Design 27

CIRCULAR PITCH

N
DP = —
PD
where
N = number of teeth
PD = pitch diameter, in.
The circumference of the pitch diameter divided by the number of teeth is called
the circular pitch:

PD
CP = tt • —
N
28 Gear Drive Systems

The circumference of the base diameter divided by the number of teeth is called
the base pitch PB:

The relationship between circular pitch and diametral pitch is

CP • DP = 7T
The area between the pitch diameter and the outside diameter is called the
addendum and on a standard gear tooth is 1.0/DP. The area between the outside
diameter and the root diameter is the whole depth, which is the addendum plus
dedendum. The whole depth of a standard gear tooth is generally 2.25 to 2.4
divided by the diametral pitch.

GEAR TOOTH GENERATION

Let us look at the generation of a gear tooth. Figure 2.6 shows a straight-sided
cutting tool, such as that used in the hobbing process, generating an involute
tooth. By “generating” it is meant that the tool is cutting a conjugate form.
Such a straight-sided tooth is sometimes referred to as a rack. As the tool
traverses and the work rotates, an involute is generated on the gear tooth flank
and a trochoid in the root fillet, as shown in Figure 2.7.
Figure 2.8 is a closer look at a hob tooth. This is a hob of pressure angle (j>
and diametral pitch 7r/(TH + TP). It is capable of cutting a whole family of gears
with the same pressure angle and diametral pitch. Such tools are standard, as,
for instance, a 20° pressure angle and 8 pitch (diametral pitch) hob, and are
easily obtainable. Gears cut by this hob will be capable of meshing with one
another. The distance (B + R) on the hob tooth is equal to the dedendum of the

CUTTING TOOL

Figure 2.6 Generating a gear tooth.

 Gear Tooth Design

Y j AXIS

Figure 2.8 Hob geometry.

30 Gear Drive Systems

Figure 2.9 Plot of involute and trochoid.

generated gear tooth. R is the hob tip radius with its center at point Z. TP, the
hob tooth space, is equal to the tooth thickness of the generated gear at the gear
pitch diameter. TH is the hob tooth thickness. When the hob traverses a distance
(TP + TH), the gear rotates through an angle (TP + TH)/RG, where RG is the
gear pitch radius. (TP + TH) is the circular pitch of the gear.
The involute and trochoid can be plotted on a Cartesian coordinate system
emanating at the center of the gear with the Y axis going through the center of
the gear tooth, as shown in Figure 2.7. Figure 2.9 is a plot of the profile of a
hobbed 20° pressure angle tooth on a Cartesian coordinate system. In the
following paragraphs equations are developed to generate this plot. These equa­
tions are easily programmed and the coordinates can be plotted automatically.
First, the involute coordinates will be obtained. Previously, it was shown
that if the base circle radius is known, the involute angle 6 and the radius to the
curve can be found for any assumed pressure angle. To find the coordinates with
respect to the center of the tooth, the tooth thickness at any radius must be
Gear Tooth Design 31

known. Let us start at the pitch diameter with a pressure angle 0 i , a pitch radius
R i , and a circular tooth thickness C TTi. The involute angle is
6 j = tan 0 i - 0 i
Referring to Figure 2.10, it can be seen that the angle A is
V2CTT 1
a = e x + ------------
Ri
To find the circular tooth thickness at any other radius R2, we use
V iC T T !
B = A - 92 = 9i + ----------- - 02
Ri

02 = COS
-1 ------
rb
R-2

6 2 = tan 02 - 02
32 Gear Drive Systems

/ V2 CTT ! \
c t t 2 = 2R2 y —- — + d x - d 2 J

To find the X and Y coordinates of the involute at any radius R2, we use
X = R2 sin B
Y - R2 c o s B
Following are the steps that would be used in a computer routine to cal­
culate the involute profile:

Ri = pitch radius

0i = pressure angle at pitch radius

CTTj = circular tooth thickness at pitch radius

rB = base radius

= tan (j>i - 0!

02 = 0.0

d2 = tan 02 —(f>2

rb
r2 =
COS 02
/ VzCTT! \
ctt2 = 2R2( - ^ ---- + 0 ! - 02 j

1/2CTT2
B =
R2
X = R2 sin B
Y = R2 cos B
Write (X, Y)
If (CTT2 • LE • 0.005) go to 2

02 = 02 + —
180
Go to 1
2 Continue
Gear Tooth Design 33

The routine starts at the base circle, where the pressure angle is 0.0, and calcu­
lates coordinates at pressure angle intervals of 7r/180 rad until it reaches a point
near the tip of the tooth where the thickness is less than 0.005 in.
In the hobbing process depicted in Figure 2.6, the point on the corner of
the tool generates the shape of the root fillet. This shape, which is bounded by
the root diameter and the involute, is called a trochoid. In Figure 2.7 the com­
plete curve is shown as a loop on the XT and YT axes. Only that portion of the
curve that lies between the involute and root diameter is of interest. To calculate
the trochoid coordinates, the following procedure is used:
1. Calculate the trochoid generated by the center of the hob tip radius, point Z
on Figure 2.8.
2. Find the normal to this trochoid at any point in order to add the radius R.
This step would not be necessary if the hob had a sharp corner since the
corner point would cut a trochoid, but practical hobs have rounded tips.
3. The trochoid coordinates calculated will be with respect to the center of the
trochoid (Figure 2.7, XT and YT axes). To find the coordinates on the
desired X-Y system through the center of the tooth they must be shifted
through the angle W (Figure 2.7).
When the hob traverses a distance (TH + TP) (Figure 2.8), the gear rotates
through an angle (TH + TP)/RG; therefore, the angle (W + V) between the center
of the gear tooth and the center of the tooth space is
Vi(TH + TP)
W+ V =
%
where Rq is the gear pitch radius in inches. Angle V in Figure 2.7 can be calcu­
lated as follows:

where
L = distance between the center of the hob tooth and point Z on
Figure 2.8, in.

T
L = TH
— - B*
tan A R
0 --------
2 cos 0

and

... _ %(TH + TP)-L

w — --------------------- ---------
%
34 Gear Drive Systems

Figure 2.11 shows the trochoid generated by point Z at its starting point and
after the hob has moved a distance R q *E and the gear has rotated through an
angle E. The coordinates are:

x z = Rz sin (T - E)
= Rz (sin T cos E - cos T sin E)

Yz = Rz cos (T - E)

= Rz (cos T cos E + sin T sin E)

Gear Tooth Design 35

where
rg - b
cos T
Rz

. xT = ---------
sm rg - e

Rz
Therefore,

XZ = (R G • E ) cos E - ( R G - B) sin E (2 ^
Yz = (R g - B) cos E + (R g • E) sin E

Figure 2.12 shows how to calculate the actual trochoid coordinates, adding the
hob tip radius R to the trochoid generated by point Z:

XT = Xz + R cos A ^ 2)
Y^ = Yz —R sin A
A is the angle formed by a line normal to the trochoid generated by point Z and
the Yt axis and can be found as follows:
dXz
tan A = -----
dYz
To find dXz /dYz :
dX7
= -(R g • E) sin E + RG cos E - RG cos E + B cos E
dE
dY7
= -R g sin E + B sin E + (RG • E) cos E + RG sin E
dE
and dXz /dYz is
dXz -(Rg • E) sin E + B cos E
dYz B sin E + (RG • E) cos E
Finally, to obtain the trochoid coordinates with respect to the system through
the gear tooth center, refer to Figure 2.7:
XT + X cos W
sin W =
Y

Y x - X sin W
cos W
36 Gear Drive Systems

Figure 2.12 Trochoid coordinates.

Xx + X cos W _ Yx - X sin W
sin W cos W
Xx cos W + X cos2 W = Yx sin W - X sin2 W (2.3)
X = Yx sin W - Xx cos W
Y = Yx cos W + Xx sin W
Let us review the procedure for plotting the trochoid coordinates:
1. Choose angles E at random and calculate the trochoid coordinates of point Z
using Eqs. (2.1).
Gear Tooth Design 37

2. Calculate the angle A at each point and define the actual trochoid coor­
dinates using Eqs. (2.2).
3. Convert to the Cartesian coordinate system through the tooth center using
Eqs. (2.3).
Following is a computer routine to carry out the process described above,
starting with angle E equal to 0 rad (where angle A is 90° or n/2 rad) and ending
with E equal to 25tt/ 180 rad.

Counter = 25 8 ——
180

L = ^TH - B tan </>- —

COS 0

w = (™ + Tp)/2 - L
%
E = 0.0
dXz
1 = -(R g ° E) sin E + B c o s E
dE
dYz
= B sin E + (Rn • E) cos E
dE
Xz = (R g 9 E) cos E - (RG - B) sin E
Yz = (R g - B) cos E + (R g • E) sin E
If (E • LE 8 0.0) go to 2

Go to 3

2 A = -
2

3 XT = Xz + R cos A

Yt = Yz - R sin A
X = Yx sin W - Xx cos W
Y = Yx cos W + Xx sin W
If (E • GE • counter), go to 4
38 Gear Drive Systems

Write X, Y

E = E+° ^
180
Go to 1
4 Continue
The previous discussion was concerned with a straight-sided cutter such as a hob.
Another type of cutting tool is in the form of a gear tooth and is called a shaper
cutter. The shaper cutter meshes with the work and generates a mating gear.
Figure 2.13 shows a shaper cutter tooth with the profile in the form of an
involute and with a rounded tip edge of radius R.
Rc = cutter pitch radius, in.
RBc = cutter base circle radius, in.
R = round edge radius, in.
Gear Tooth Design 39

R0 c = cutter outside radius, in.

Z = center of round edge
As the cutter rolls through angle C the gear it is cutting will roll through an angle
C • Rc /Rg , where RG is the gear pitch radius. Referring back to Figure 2.7, the
angle W between the center of the trochoid and the center of the gear tooth is
%CTTg / C ‘ Rr
W = ---------- + 1 ^
Rg \ Rg

where CTTG is the circular tooth thickness of the gear at the pitch diameter, in
inches. From Figure 2.13 angle C is found as follows:
C ~ 0 r + M - 0£

M = Inv 0R - Inv 0

r bc
02 = cos 1
R qc - R
PZ = Rbg tan 0Z
PZ + R
0R = tan" — -----
KBC

where 0 is the pressure angle at the pitch diameter.

Figure 2.14 shows how the coordinates of the trochoid of point Z, the
center of the shaper cutter round edge radius, are arrived at:

XZ = R Z sin (T - Eg )

Yz = Rz cos (T - Eg )
sin Ec
sinT - (R qc ” R)
RZ

Rz + (Rc + R g )2 - (R0C - R )2
cos T = ---------------------------------------------
2(R z )(R c + Rg )

Rz = (Roc - R)2 + (RC + Rg )2 " 2(Roc “ R)(Rc + rg ) cos eg

Therefore,

xz = (Ro c - R) sin (Ec + Eg) - (Rc + r g ) sin e g

Yz = (R^ + R g) cos e g ” (Roc " R) cos (Ec + e g )

40 Gear Drive Systems

To find the trochoid coordinates with respect to the center of the gear tooth,
the same procedure is followed as was used for the hob-type cutter:
1. From Figure 2.12 find the XT and YT coordinates differentiating the fore­
going equations to calculate tan A using Eqs. (2.2).
2. From Figure 2.7, shift to the X, Y coordinate system by using Eqs. (2.3).

GEAR TEETH SN ACTION

Figure 2.15 shows a gear mesh with the driving pinion tooth on the left just
coming into mesh at point T and the two teeth on the right meshing at point S.
Gear Tooth Design 41

Notice that contact starts at point T where the outside diameter of the gear
crosses the line of action and ends where the outside diameter of the pinion
crosses the line of action, point R.
Z is the length of the line of action. In other words, a tooth will be in con­
tact from point T to point R. PB is the base pitch, the distance from one
involute to the next along a radius of curvature. It was shown earlier that

where
BD = base diameter, in.
N = number of teeth
Point T, where contact initiates, is called the lowest point of contact on the
pinion tooth and also the highest point of contact on the gear tooth. Similarly,
point R is the highest point of contact on the pinion tooth and the lowest point
of contact on the gear tooth. Point S is the highest point of single tooth contact
on the pinion and the lowest point of single tooth contact on the gear. In other
words, if one imagines the gears in Figure 2.15 to begin rotating, just prior to

Figure 2.15 Gear tooth action.

42 Gear Drive Systems

meshing at point T a single pair of teeth was carrying the load. As the gears
continue turning and the pinion tooth on the right moves from point S to R,
two pairs of teeth are carrying the load and after point R a single pair again
carries the load until the next two teeth mesh at point T.
Thus it can be seen that for some period of time one tooth mesh carries
the load and for another period of time two tooth meshes share the load. A
measure of the percentage of time two meshes share the load is the profile
contact ratio Mp. For instance, a profile contact ratio of 1.0 would mean that
one tooth is in contact 100% of the time. A contact ratio of 1.6 means that two
pairs of teeth are in contact 60% of the time and one pair carries the load 40%
of the time. Contact ratios for conventional gearing are generally in the range
1.4 to 1.6.
Let us now derive the profile contact ratio in terms of parameters easily
obtainable:

e tr

where
Etr = degrees of roll to traverse the length of the line of action from
point T to R
N = number of teeth
This equation may not be obvious, but it can be understood if it is remembered
that from one tooth to another the base circle must roll 360°/N. Thus if the base
circle rolls 360°/N while going from point T to R, the profile contact ratio is
1.0. If the base circle rolls more than 360°/N going from T to R, the profile
contact ratio is greater than 1, indicating that for some percentage of time two
pairs of teeth are in contact.
Et r , the total degrees of roll, is equal to the degrees of roll to the pinion
tooth tip minus the degrees of roll to the pinion lowest point of contact:
etr = e odp - E tifp

where
E0DP = degrees of roll to pinion outside diameter
Etifp = degrees of roll to true involute form diameter on the pinion
(lowest point of contact) (point T on Figure 2.15)
Figure 2.16 shows how the degrees of roll to the pinion outside diameter is cal­
culated:
Gear Tooth Design 43

Figure 2.16 Degrees of roll to pinion outside diameter.

/ 1§0 \ \ / RODP ■ RBP /180

Eqdp = ta n 0 ODp( r j - - Rbp [ n

Figure 2.17 shows how the degrees of roll to the pinion form diameter is cal­
culated:

A80\ _ V^R j i f p * r bp /l 80
F jif p _ tan 0xiFP
7T J Rgp \ ^
It is more convenient to express ETIFP in terms of the gear outside radius and
base radius:
XX — C sin 0pj)

V^Rxifp - Rbp = C sin 0 pd - %/r odg ~ r bg

44 Gear Drive Systems

Figure 2.17 Degrees of roll to pinion true involute form diameter.

Therefore,

„ _ C sin 0pD
^ - \/R o d g ” RB
^BG
^TIFP
RBP f? )
and

\/R odp " r BP ~ ^ sin ^PD + >/r ODG ” r BG 180N

MP = -----------------------------------------------------------------------
RBP 71*3 60

where
N = number of pinion teeth
Rgp = %PDp cos 0pD
Gear Tooth Design 45

Therefore,

\ / r ODP " r BP “ C sin ^PD + \/R^)DG “ RBG

Mp = ----------------------------------------------------------------
CP cos 0pj)

Another way of expressing the profile contact ratio is

CP COS 0 p j ) = Pg
because

COS 0ap r \ - BD
— -----
l-U pD

and

CP-*™
N
Therefore,

nCP
v COS 0rkp D = 7T —
BD -- D
Pg

and

\i R ODP " R BP " C sin ^PD + %/R ODG “ R BG ” Z

Therefore,

Mp = A
bb
To calculate the degrees of roll to the highest point of single tooth contact on
the pinion, consider Figure 2.18. The distance XS is the sum of XT from Figure
2.18 and TS from Figure 2.15. The distance XT can be calculated using Figure 2.17:

XT = V r I jfp - Rg
BP

From Figure 2.15 it is seen that the distance TS is a base pitch; therefore,

XS = \ / r tifp ~ r bp + Pb
46 Gear Drive Systems

Figure 2.18 Degrees of roll to pinion highest single tooth contact diameter.

v —+ j. / 180\ _ XS /180 \
E HSTCP “ ta n “— V 1 /

= ^ RTIFP " R|p + EB / 180\

r bp I 77 /
or

e h s tc p = e tif p + —
The involute curve changes very rapidly near the base diameter and more slowly
at sections farther away from the base circle. This is illustrated in Figure 2.19,
which shows the difference in length along the involute curve for equal incre­
ments taken on the base circle. The distance XY is far less than YZ. Since the
Gear Tooth Design 47

involute is so sensitive near the base circle, the lowest point of contact on a gear
tooth should be located well away from the base circle. As a rule of thumb the
lowest point of contact on a gear tooth should be at least 9° of roll.

ROLLING AND SLIDING VELOCITIES

When involute gear teeth mesh, the action is not pure rolling as it would be when
two friction disks are in contact, but a combination of rolling and sliding. Figure
2.20 shows a gear mesh with two base circles of equal size and the teeth meshing
at the pitch point. Radii of curvature are drawn to the involutes from equal
angular intervals on the base circle. It can be seen that arc XY on gear 2 will
mesh with arc AB on gear 1 and that AB is longer than XY; therefore, the two
profiles must slide past one another to make up the difference in length. The
sliding velocity, which is usually expressed in feet per minute, at any point is
calculated as follows:
48 Gear Drive Systems

W jR d - W2Rc 2
V c “ -------------------------------
b 12
where
Vs = sliding velocity, fpm
Wj = angular velocity of gear 1, rad/min
W2 = angular velocity of gear 2, rad/min
RC1 = radius of curvature of gear 1, in.
RC2 - radius o f curvature o f gear 2, in.

From Figure 2.20 it can be seen that when point A on gear 1 and point Y on
gear 2 mesh, RC1 will be larger than RC2 and since Wj_ = W2, Vg will be a posi­
tive number. As the meshing point nears the pitch point the difference in the
radii of curvature lessens until at the pitch point the radii of curvature are equal
and Vg is 0. When point A on gear 2 meshes with point Y on gear 1, RC1 will
be smaller than RC2 and Vs will be negative. The significance of this is that as
Gear Tooth Design 49

the mesh goes through the pitch point the direction of sliding changes. There is
always pure rolling at the pitch point. If the base circles were of unequal size at
the pitch point, W1RC1 would still equal W2RC2 since

W^ R2 Rc j I sin 0
W2 Rj RG2/sin 0
where
R l = pitch radius of gearvl , in.
R2 = pitch radius of gear 2, in.
0 = pressure angle, deg
Sliding velocity is significant in that it affects the amount of heat generated in
the gear mesh. Also, the fact that gear teeth undergo sliding as well as rolling
must be appreciated.
Another significant velocity term is the sum or entraining velocity:
VE = W iR cj + W2Rc 2
where VE is the sum velocity in ips. The sum or entraining velocity is a measure
of how quickly oil is being dragged into the conjunction between the two gear
members.
The parameter generally used when expressing the speed of a gearset is the
pitch line velocity:
WPRp(60) _ Wg Rg (60)
vx 12 12
rrDpnp 7rDGnG
12~ ~ 12

where
vT = pitch line velocity, fpm
Wp = pinion angular velocity, rad/sec
wG = gear angular velocity, rad/sec
Rp = pinion pitch radius, in.
Rg = gear pitch radius, in.
dp = pinion pitch diameter, in.
d G = gear pitch diameter, in.
n p = pinion rpm
nG =gear rpm
The pitch line velocity is a measure of the tangential or peripheral velocity of a
gearset and a better indication of speed than the rpm. For instance, a 1-in. pitch
50 Gear Drive Systems

diameter gear operating at 10,000 rpm has the same pitch line velocity as a
10-in. pitch diameter gear operating at 1000 rpm. Two meshing gears always
have the same pitch line velocity.
American Gear Manufacturers Association (AGMA) Standards for enclosed
drives consider units with pitch line velocities of 5000 fpm or more high speed.
Gear units have been operated at pitch line velocities up to 50,000 fpm minute
but applications over approximately 20,000 fpm require extremely careful
analysis concerning lubrication, cooling, and centrifugal effects.

HELICAL GEARS

Figure 1.13 illustrates the difference between spur and helical gears. The tooth
contact on spur gears is a straight line across the tooth and at any time either
one or two teeth are in contact. The helical gear contact, because the teeth are at
an angle to the axis of rotation, is a series of oblique lines with several teeth in
contact simultaneously and the total length of contact varies as the teeth go
through the mesh.
To understand the nature of the helical tooth, consider a base cylinder
with a series of strings wrapped around it as shown in Figure 2.21. The start of
each string is offset such that a line joining the string starts is at an angle to
the axis of rotation of the cylinder. The ends of each string when held taut and
unwrapped from the base cylinder will define involutes and the surface defined
by the string ends will be a helical involute gear tooth.

INVO LUTE
BASE RADIUS
Gear Tooth Design 51

In one rotation of the base cylinder an axial length L of strings will be

unwrapped. L is defined as the lead

where
L = lead, in.
RB = base radius, in.
0 g = base helix angle, deg
The helix angle along the tooth profile varies with the radius but the lead is a
constant. Once the base radius and base helix angle are defined, the lead can be
calculated and the helix angle at any radius R is known:

where 0 R is the helix angle at radius R, in degrees.

Transverse and Normal Planes

Figure 2.22 shows the relationship between the transverse and normal planes of
a helical gear. The transverse plane ABCD is the plane of rotation, while the
normal plane ABE is at right angles to the tooth. The normal and transverse
planes are displaced from each other through the helix angle 0. Following are
the relationships between the normal and transverse pressure angles at any radius
on the tooth illustrated at point B on Figure 2.22.
tan 0N = AB
AE
AB
tan 0T
AD
AD
COS 0
AE
AD tan 0T = AE tan 0N
Therefore,
cos 0 tan 0T = tan 0N
Figure 2.23 shows a helical gear rotating about the axis XX. The teeth are
inclided with relation to the axis of rotation the helix angle 0. Usually, the helix
angle at the operating pitch diameter is referred to. When the inclination of the
Gear Drive Systems
Gear Tooth Design 53

teeth is off to the right as shown in Figure 2.23, the gear helix is designated as
right hand. When two external helical gears mesh, one must be right hand and
the other left hand. When an external gear meshes with an internal gear they will
both have the same hand of helix.
The circular pitch in the transverse plane CPT has the following relation­
ship with the circular pitch CPN in the normal plane:
, CPN
COS w = ---------
CPX
The circular tooth thicknesses have the same relationship:
CTTN
cos \p = -------
CTTT
where
CTTN = normal circular tooth thickness, in.
CTTT = transverse circular tooth thickness, in.
The normal and transverse diametral pitches have the relationship
DPn c o s 4/ = DPX
The distance along the tooth axis from one tooth to another is called the axial
pitch PA, as shown in Figure 2.23. The ratio of face width F to axial pitch is
called the face contact ratio or the helical overlap and is a measure analogous to
the profile contact ratio for spur gears. The face contact ratio, designated as
MF , is

CPN COS \jj 77

PA = — - CPX --------- = — .-
sin \p sin ijj DPX tan \p
Therefore,

MF = F(DPX) ^ - ^ -
7T
The total contact of a helical gear mesh is therefore some combination of the
profile and face contact ratios. Sometimes the sum of the two is called the total
contact ratio and used as a measure of how much contact is achieved in a tooth
mesh.
The actual total length of contact at any instant in a helical mesh is the
sum of the length of the oblique lines of contact on each tooth in mesh and
54 Gear Drive Systems

varies as the teeth go through mesh. A method of calculating the minimum and
maximum length of the lines of contact was derived by E. J. Wellauer and pre­
sented to the Industrial Mathematics Society. The paper was entitled “The
Nature of the Helical Gear Oblique Contact Line,” and a small portion based on
the original article is given below.
Ka and na are the whole number and fractional portion, respectively, of
the face contact ratio. For example, if Mp = F/PA = 4.85, then Ka = 4.0
and na = 0.85. Kr and nr are the whole number and fractional portion,
respectively, of the profile contact ratio. For example, if Mp = Z/PB
= 1.32, then Kr = 1.0 and nr = 0.32.
If (1 - nr)/na ^ 1 , then

(Z-F/Pb ) - nrnaPA
^min
COS \ p B

where
Lmin = minimum total length of the oblique lines of contact, in.
Z = length of the line of action, in. (Figure 2.15)
PB = base pitch, in.
PA = axial pitch, in.
i//B = base helix angle, deg
If (1 - nr)/na < 1, then
_ (Z-F/PB) - ( l - n a) ( l - n r)PA
min cos <//B
An approximation used for calculating Lmin in several AGMA Standards is
0.95 (Z)F
^min
PB COS ip B

To calculate the maximum total length of the lines of contact:

If nr < na, then
_ (Z-F/Pb ) + nr(l - n a)PA
^max cos i//B

If nr > na, then

(Z-F/P b ) + n a(l - n r)PA
^max
COS l//B
Gear Tooth Design 55

INTERNAL GEARS

The involute form of internal gears (sometimes called ring gears) is the same as
for external gears. The difference between the two lies in the fact that internal
gears contact on the concave side of the involute rather than the convex. Also,
the root diameter of an internal gear tooth is the largest diameter and the tips
of the teeth are at the inside diameter, which is the smallest. Figure 2.24 shows
internal gear tooth geometry. At any radius R, with pressure angle 0 and base
radius RB, all the involutometry calculations will be essentially the same as those
previously shown for external gear teeth.

Figure 2.24 Internal gear geometry.

56 Gear Drive Systems

Figure 2.25 Internal gear mesh action.

To illustrate internal gear mesh calculations, let us derive the profile con­
tact ratio for the situation shown in Figure 2.25 for an external pinion driving
an internal gear. The start of contact is at point R, where the internal gear inside
diameter crosses the line of action. Contact ends at point T, where the pinion
outside diameter crosses the line of action. The total length of contact is the
distance from R to T or Z as shown on Figure 2.25. One pair of teeth is meshing
at point S on the figure and an adjacent pair at point T; therefore, the distance
ST is a base pitch PB. The profile contact ratio is

e tr
MP =
360°/N
Gear Tooth Design 57

where
Etr = degrees of roll to traverse the length of the line of action from
point R to T
N = number of pinion teeth
Ej r , the total degrees of roll, is equal to the degrees of roll to the pinion outside
diameter minus the degrees of roll to the lowest point of contact on the pinion
(TIF, true involute form diameter).

e tr = E ODP " e t if p

where
E0Dp = degrees of roll to the pinion outside diameter
Etifp = degrees of roll to the pinion true involute form diameter

As shown previously (Figure 2.16),

c - _ /1 8 0 \ \ / R odp • Rlp /ISO \

0DP ' ,a" * ODP j ( V )

where
R0 dp = pinion outside radius, in.
R bp = pinion base radius, in.
Figure 2.26 shows how the degrees of roll to the pinion form diameter is cal­
culated:

r; _ , . - / 180\ \ / RTIFP " RBP /180 \

e TIFP - tan 0TIFp j ------^ ----------- V 7T /

It is more convenient to express ETjpp in terms of the gear inside radius and base
radius. From Figure 2.26,
XX = C sin 0pj)
where 0PD is the pressure angle at the pitch diameter, in degrees.

\M lF P " R BP = \ / R IDG ” ^ B G “ E sin 0P D

where

R jdg = gear inside radius, in.

RfiG = §ear base radius, in.
58 Gear Drive Systems

Figure 2.26 Internal gear mesh degrees of roll to pinion form diameter.

Therefore,

/R fc c " R BG ~ C sin ^PD

e t if p
r bp

and

\/R q d p “ r bp + E sin 0PD " \ / R rDG ~ r bg 180N

Mp
R gp it* 3 60
Rbp = %PDp cos 0pD
N 1
- 7r
PDP CP
Gear Tooth Design 59

where CP is the circular pitch in inches. Therefore,

% /R o d p " r bp + C sin ^PD “ A iD G ~ r bg

Mp = -------------------------------------------------------------
CP COS 0ppj

MEASUREMENT OVER BALLS OR WIRES

This subject is presented at this point not only because it is an important mea­
surement in the manufacture of gear teeth but because it is a good illustration of
the application of involutometry in the analysis of gear tooth geometry.
When cutting or grinding a gear tooth the machine operator will check the
tooth thickness to determine when sufficient stock has been removed from the
flank to bring the tooth to the required size. The drawing requirement may call
for a tooth thickness at a given diameter. This is a difficult measurement to
make directly; therefore, quite often an indirect measurement is used. Balls or
wires (sometimes called pins) of a known diameter are placed in 180° opposite
tooth spaces on the gear and an accurate micrometer measurement over the balls
or wires is made.
The equations for calculating measurement over balls or wires will be
derived first for an external spur gear with an even number of teeth. In this case
two opposite tooth spaces will be in line. The analysis will then be extended to
gears with odd numbers of teeth where the opposite tooth spaces are not in line
and then internal gears and helical gears will be addressed.
Referring to Figure 2.27, for a spur gear with an even number of teeth the
measurement over wires (MOW) is
MOW = 2R 2 + 2X
where
R2 = radius to the center of the wire, in.
X = wire radius, in.
We are going to calculate the MOW for a gear with a known circular tooth thick­
ness T at a known radius R x.
Rx cos 0x
i—
R2 = —cos <p2
where

0x = pressure angle at the radius Rx, deg

02 = pressure angle at radius R2, deg
60 Gear Drive Systems

The problem now is to calculate the angle 02 • To accomplish this, an imaginary

involute is drawn through the center of the wire as shown by the dashed profile
in Figure 2.27. From this construction we can see that angle D is the involute
of 02.

D = Inv 02 = B + C + A - —
N
where

N = number of gear teeth

C = inv 0 j , which can be calculated knowing 0 X; Inv 0 X= tan 0i - 0i
Gear Tooth Design

Figure 2.28 Measurement over wires for gear with odd number of teeth.

A = ——
2Rj
X
B = — , where RB = gear base radius
Rn
62 Gear Drive Systems

B = X/R b because the circular distance between two involutes on the base circle
is equal to the distance between normals to the involutes. In other words, the
wire radius X is equal to the base pitch between the imaginary involute and the
adjacent tooth involute. Knowing the involute of 02 >the angle 02 can be calcu­
lated using the involute tables in the Appendix and then R2 and the measure­
ment over wires can be calculated.
When the gear has an odd number of teeth the situation is as shown in
Figure 2.28 and the fact that the tooth spaces do not line up must be compen­
sated for mathematically. In Figure 2.28:

Figure 2.29 Measurement between wires of internal spur gear.

Gear Tooth Design 63

90
A A = AB =
~ N~
and

MOW = 2|r 2 cos| ^ + x j

Figure 2.29 illustrates the analysis for the measurement between wires for an
integral gear with an even number of teeth:
MBW = 2R2 - 2X
COS 0 !
R2 - Ri
COS 0 2

Inv <t>2 = D = — - B- A + C
N
Again an imaginary involute is drawn through the center of the wire and the
involute 02 is calculated from which the angle 02 and R2 can be derived. For
odd numbers of teeth,

MBW = 2^R2 cos^ ~ j ' X]

Figure 2.30 illustrates a ball placed between two helical gear teeth. When mea­
suring helical gears balls should be used rather than wires since the wires will not
seat properly in the helices. The balls will contact the gear teeth in the normal
plane, but the measurement over balls calculation must be made in the transverse
plane. Figure 2.30 shows how the projection of the ball is mathematically
shifted into the transverse plane and the equation for the involute 02 is
t j X , j , , T tr
Inv 02 = ------------- + Inv 0x +------ -------
R b cos 0 B 2RX N
for external gears and

Inv 02 = - -----—------- + Inv0 ! ----- — + —

RB cos 0 B 2Ri N
for internal gears. It should be noted that the circular tooth thickness T is in the
transverse plane. If a normal tooth thickness is given, it should be shifted to the
transverse plane: T = TN/cos 0.
To sum up, the measurement over or between balls or wires for gears with
even numbers of teeth is
 64
Gear Drive Systems

Figure 2.30 Measurement over balls of external helical gear.

Gear Tooth Design 65

MOW = 2(R2 + X) for external gears

MBW = 2(R2 - X) for internal gears
For gears with odd numbers of teeth,

MOW
90
/on
■4
^
for external gears

MBW = 2 R2 cos for internal gears

For all gears,
( i ■-)]
cos 0!
R2 = Ri
cos 02
To find cos 02,

Inv d
X + Inv (
tr
RB cos 0 g 2RX N
for external gears and

Inv 02 = -
X + Inv < + *
Rg cos 0 B 2Ri N
for internal gears. The cosine 02 is found from Inv 02 using involute tables. For
spur gears, cos 0 B = 0.0.
When choosing ball or wire size a good estimate for the diameter is
1.728
D BALL - 2X -
DPiN
where
X = ball or wire radius, in.
DPn = normal diametral pitch

Let us work through an example to illustrate the calculation for measure­

ment over balls. Assume an external helical gear with the following dimensions:

Number of teeth 38
Normal diametral pitch 15.868103
Normal pressure angle 20.0°
Helix angle at pitch diameter 18.0°
Normal circular tooth thickness 0.0952
at pitch diameter, in.
66 Gear Drive Systems

The transverse diametral pitch is

DPX = 15.868103(cos 18.0) = 15.091463
The transverse pressure angle is

</>x = tan-1 tan 20-0 = 20.941896

cos 18.0
The pitch diameter is
oo
PD = ---- ~ ------- = 2.517980
15.091463
The base diameter is
BD = 2.517980(cos 20.941896) = 2.351651

The lead is

L = < 2:S1-79-g°) = 24.345915

tan 18
The base helix angle is
, M -, /n(2.351651)\ ,
\ 24.345915 /
The involute of the transverse pressure angle is

Inv <j> = tan 20.941896 - 20.941896| — \ = 0.017196

W
The transverse circular tooth thickness is
0 0952
CTTT = — - = 0.100099
cos 18.0
For a ball diameter of 0.125,

In v*, - ___________________________ t 0.017196 + a l0 0 0 "

(2.351651/2) cos 16.880766 2.517980

- — = 0.029824
38
Using the involute tables (Appendix) yields

02 = 24.9599°
Gear Tooth Design 67

and
(2.517980/2) cos 20.941896 _ 2.593914
R«
2 cos 24.9599 2
The measurement over balls for 0.12 5-in .-diameter balls is
1.125 \
MOB = 2 I 2-593914 + | = 2.7189
2 /

Measurement of Tooth Thickness by Calipers

Tooth thickness can be checked by measuring across several teeth with vernier
calipers as shown on Figure 2.31. The calipers contact the teeth at points X and
the line XX is tangent to the base circle. The arc AB along the base circle is equal
to the length XX:

2?rS
+i 2 T
Inv 0
N /
where
M = caliper measurement, in.
Rb = base circle radius, in.
T = given transverse tooth thickness at a radius R, in.
R = given radius, in.
0 = given transverse pressure angle at radius R, in.
S = number of tooth spaces between the contacting profiles
N = number of teeth in the gear

Figure 2.31 Tooth thickness measurement by vernier calipers.

68 Gear Drive Systems

For helical gears,

M = R t> c o s \jjy y [— + + 2 Inv 0

B B \R N
where 0 B is the base helix angle in degrees.

Center Distance and Tooth Thickness

The center distance of a pair of meshing spur or helical gears is established by

the location of the centers of the bearing bores locating the gear shafts. For the
gears to mesh properly at a given center distance, the tooth thicknesses must be
chosen such that the teeth will not bind under all operating conditions. The
following variations in center distance must be taken into account when design­
ing a gear set:
1. The center distance will vary due to tolerances in the bearing housings.
2. Clearance in the supporting bearings will affect the operating center distance.
3. Temperature variations during operation will change the operating center
distance. At a minimum, expansion of the gear teeth must be taken into
account. If dissimilar materials with varying coefficients of expansion are
used in the gearbox, their thermal growths must be analyzed.
In order to accommodate all these variables, backlash is designed into the
gear mesh. Backlash can be defined as the circular pitch minus the sum of the
circular tooth thicknesses:
BL = CP - (TP + Tg)
In most cases excessive backlash will not be harmful and is much more desirable
than too little backlash, which can result in tight meshing and binding of the
gears. In very high speed helical gearing it is important to have sufficient back­
lash to allow the air-oil mixture being pumped between the teeth to exit the
mesh without becoming excessively churned and heated. The amount of back­
lash designed into a gear mesh will vary with the diametral pitch of the teeth.
Following is a table of suggested backlash versus pitch.

Diametral pitch Nominal backlash, (in.)

6 0.015
8 and 10 0.010
12 and 14 0.009
16 0.008
18 0.007
20 0.006
Gear Tooth Design 69

Excessive backlash may be detrimental if the transmitted load varies to the

extent that the tooth can become unloaded and contact on the normally
unloaded face. In this case the more backlash there is in the mesh, the greater
freedom the teeth will have to rattle around and the greater dynamic load
generated. There are also cases where gears are used as positioning devices and
backlash is detrimental. In such designs special techniques such as adjustable
center distance are used to achieve low backlash.
Center distance can be expressed mathematically in various ways:
PDp + PDq
C = --------------- for external gears
2
PDg - PDp
C = --------------- for internal gears
2
where
C = center distance, in.
PDp = pinion pitch diameter, in.
PDq = gear pitch diameter, in.

Since PDp/Np = PDG/NG =: DPX,

NP + Ng
C = ------------- for external gears
2DPX
Ng - NP
C = ------------------for internal gears
2DPX
where
Np = number of pinion teeth
Ng = number of gear teeth
DPX = transverse diametral pitch
If R is the gear ratio PDG/PDp,
PDP(1 + R)
C = --------------- for external gears
2
P D p (R - 1)
C = -------------- for internal gears
2
Let us look at a so-called standard spur gearset, where a standard diametral
pitch cutter is used:
70 Gear Drive Systems

DPX = 10.0
0T1 = 20 °
NP = 20
Ng = 30

Cl = ^ = 2 . 5
2( 10)
If the backlash is 0.010,

TP1 = t G1 = — - — = — ----------------- = 0.152080

F1 01 2 2 ( 10)2 2
where
TP1 = pinion transverse circular tooth thickness atthe pitch radius RP1
Tg i = gear transverse circular tooth thickness at the pitch radius RG1
Let us calculate at what center distance C2 the mesh will have zero back­
lash. At this point of tight mesh or binding,
27fRp9 27 t R g ?
CP = t P2 + t G2 = — ^
NP Ng
where
CP = transverse circular pitch, in.
TP2 = pinion transverse circular tooth thickness atthe tight mesh pitch
radius Rp2, in.
Tq2 = gear transverse circular tooth thickness at the tight mesh pitch
radius RG2, in.
As shown previously:

TPo = 2 R p J — + Inv 0! - Inv 02

\2R pi

TG2 = 2RG2( —— + Inv 0! - Inv 02

\2 R o i i
Also,

N° = Rg i =

NP Rpi Rp2
Combining the four equations above, we have
Gear Tooth Design 71

2RP1(NP + NG)

COS 0 2 R pi R q i

COS 0 i Rp2 R q 2

and the tight mesh center distance C2 = cos 0i/cos 02. For the example
above,
Inv 02 = 0.012904
02 = 19.0910°
C2 = 2.4860
For internal gears the equation for TG2 is

and

In this case TG1 and TG2 are circular tooth thicknesses of the internal gear.
The situation is somewhat more complicated for helical gears since stan­
dard diametral pitch, pressure angle, and tooth thickness are defined in the
normal plane, yet the calculations are carried out in the transverse plane. To
work helical gear problems, all normal values must be transferred to the trans­
verse plane prior to calculating. Quite often a series of gears is designed to
achieve different ratios on the same center distance. Let us look at a helical gear
example to illustrate the mathematics involved.
An electric motor operating at 3550 rpm drives a compressor. It is de­
signed to operate the compressor at two different speeds, 33,897 and 31,842
rpm. With a 296-tooth gear driving, a 31-tooth pinion will achieve 33,897 rpm
and a 33-tooth pinion will achieve 31,842 rpm. The idea is to use the same gear,
housing, bearings, and so on, for both ratios, only changing the pinion to achieve
either ratio. Also, it is desired to use 20 diametral pitch, 20° pressure angle
cutters for all gears. The center distance is chosen by stress considerations as
8.4780 in. In order to encompass both the 31- and 33-tooth pinion designs, let
us first calculate the gear geometry for a 32-tooth pinion design. The same gear
will then mesh with the 31- and 33-tooth pinions.
The transverse diametral pitch of the 296 X 32 design is

DPX = - 96 + -32 = 19.34418495

1 2(8.4780)
72 Gear Drive Systems

The gear pitch diameter is

296
PDr = ---------- ------ = 15.3017561
u 19.34418495
In order to use a standard 20 normal diametral pitch cutter the helix angle is
.
w = cos
_! -----------------
19.34418495 = 14.71320405
20
The gear lead is

= 7r(15-3°l 7S61) = 183 0672333

tan 14.71320405
For a normal pressure angle of 20° the transverse pressure angle is

<pT = ta n '1 ----- --------------- = 20.62180626°

cos 14.71320405
For a standard gearset the pinion and gear outside diameter would be set by
using a standard addendum of 1/DP; therefore, the gear outside diameter would
be

ODr = 15.3017561 + — = 15.402

° 20
When a large gear is meshing with a small pinion it is conventional to increase the
pinion addendum and decrease the gear addendum, resulting in what is com­
monly called a long and short addendum design. In the example the pinion
addendum is 0.0629 and the gear addendum is 0.0405. These values are arrived
at using two criteria:
1. The degrees of roll to the form diameter of the pinion in the 296 X 33 mesh
must be high enough to avoid undercutting of the pinion. Undercutting
occurs when the gear tooth tip describes an arc through space that would cut
through the active profile of the pinion. In other words, the trochoid gener­
ated by the gear tooth tip would interfere with the pinion involute above the
pinion form diameter.
2. The addendums are varied such that the temperature rise in the mesh due to
sliding is minimized. This subject is discussed in Chapter 3.
It should be noted that the pinion addendum is lengthened by the same
amount the gear addendum is shortened. Because of the long and short adden­
dum design standard tooth thicknesses cannot be used, since this would result in
an imbalance of bending strength between the pinion and gear, the pinion being
weakened. Assuming a backlash in the mesh of 0.006, the standard transverse
circular tooth thicknesses for pinion and gear would be
Gear Tooth Design 73

CITTp = CTTT0 = “ - “ t ------------ ! ---------------“ 06

F u 2 2 2(19.34418495) 2
= 0.078203
The optimized transverse circular tooth thicknesses for the 296 X 32 mesh ar
are:
CTTTp = 0.09072
CTTTg = 0.06569
Now let us look at the 296 X 31 tooth mesh. The transverse diametral pitch is
296 + 31
DPX = = 19.28520878
2(8.4780)
and the gear pitch diameter is

PDr = ----- — = 15.34855046

U 19.28520878
The helix angle at this diameter is

$ = tan-1 ( rcC15-34855046) ^ = 14.75623793°

\ 183.0672333 )
and the transverse pressure angle is
= eos_, / « a _20.62180626(15.3017S61) \ g 2 , 0 8 m 6 9 7 .
1 \ 15.34855046 /
Knowing the gear transverse circular tooth thickness at the 15.3017561 diameter
(0.06569), we can calculate the gear transverse circular tooth thickness at the
15.34855046 diameter:
CTTTg = 0.0480391
and to achieve a backlash of 0.006 the pinion transverse circular tooth thickness
is

CTTTp = ------------- - 0.0480391- 0.006 = 0.10886192

19.28520878
Since we cannot optimize the addendums for both the 296 X 31 and 296 X 33
tooth meshes we will use the pinion outside diameter calculated for the 296
X 32 mesh 1.780 for the 31- and 33-tooth pinions. Figures 2.32 and 2.33 are
computer printouts giving all the tooth geometry for the 296 X 31 and 296
X 33 meshes.
 74 Gear Drive Systems

AMERICAN LOHMANN rcAR

DRIVEN DRIVER
PINION EXTERNAL GEAR
Of nnnnnna
NUMBER OF TEETH 31 nVVWWvdo . wovvvvv TRANSVERSE DIAMETRAL PITCH 19.2852088
HELIX ANGLE (DEG) 14.7562380 14.7562380 TRANSVERSE PRESSURE ANGLE(DEG) 21.0811130
PITCH DIAMETER 1.6074495 15.3485505 CENTER DISTANCE 8.47800*0
00007 /9/SJ9AU9J
g/g
RELATIVE RULING SPEED (RPM) 3307! a vvWBvvv 3550.0236486 PITCH LINE VELOCITY (FPM) 14264.8838538
MESH TORQUE (IN-LBS) 316.0825442 3018.0784863 MESH RATIO 9.5483P71
A PAAAAAA a cn.mnn.nn
BENDING GEOMETRY FACTOR Vn^JufrowBv v . ZmvvvvV RELATIVE HORSEPOWER PER MESH 170.0000&tf0
BENDING STRESS (PSI) 19677.4648288 19677.4648288 EFFECTIVE FACE WIDTH 1.6250000
BENDING LIFE (HOURS) 999999.0000000 OOOOOO 7 7 7 777 . VVtruvOv STATIC TANGENTIAL LOAD (LBS) 393=27211°7
BENDING SAFETY FACTOR 2.2360604 2.2360604 DYNAMIC FACTOR 1.5349o^l
COMPRESSIVE STRESS (PSI) 84528.9062052 84528.9062052 ALIGNMENT FACTOR 1.3733637
COMPRESSIVE LIFE (HOURS) 999999.0000000 999999.0000000 MODIFIED TANGENTIAL LOAD (LBS) 829.0260354
COMPRESSIVE SAFETY FACTOR 1.6089171 1.6089171 SURFACE FINISH 20.00001^0
SLIDING VELOCITY AT TIP (FPM) 3728.8957744 -931.9806035 FLASH TEMPERATURE RISE (DEG F) 72.9815383

1 naneMnLn.
A.G.M.A. MATERIAL GRADE 1.0000000 PROFILE CONTACT RATIO 1.5640283
ALTERNATING BENDING FACTOR 1.0000000 1.0000000 FACE CONTACT RATIO 2.6274494
NUMBER OF MESHES PER REV 1.0000000 11 itFBvvmrO
anun&a&iBi MIN CONTACT LENGTH 2.5137344
BASE HELIX ANGLE (DEG) 13.8076961 13.8076961 MAXIMUM CONTACT LENGTH 2.7510025

OUTSIDE DIAMETER 1.777- 1.780 15.380-15.383 BACKLASH 0.00600*0

PITCH DIAMETER 1.6074495 15.3485505 CIRCULAR PITCH 0.1629*17
FORM DIAMETER 1.5757543 15.2158791 LEAD ERROR (IN/IN) 0.0002000

BASE DIAMETER 1.4998664 14.3213053 BASE PITCH 0.1519390

ROOT DIAMETER 1.521- 1.531 15.124-15.134 DEPTH TO POINT OF MAX SHEAR -0.0027w4

ROLL ANGLE-MAX OUTSIDE DIA 36.6168247 22.4672304

ROLL ANGLE-ROUND EDGE DIA 36.0681676 22.3643649
ROLL ANGLE-HIGH SINGLE TOOTH 30.0682702 21.7814020
ROLL ANGLE-PITCH DIA 22.0869009 22.0869009
ROLL ANGLE-LOW SINGLE TOOTH 25.0039227 21.2510143
ROLL ANGLE-FORM DIAMETER 18.4553682 20.5651859

TOP LAND THICKNESS 0.0264463 0.0347370

MAX CASE DEPTH 0.0241853 0.0261853
TRANSVERSE CIR TOOTH THICKNESS 0.1088630 0.0480387
NORMAL CIR TOOTH THICKNESS 0.1048- 0.1058 0.0460- 0.0470
NORMAL DIAMETRAL PITCH 19.9429632 19.9429632
NORMAL PRESSURE ANGLE 20.4441987 20.4441987
LEAD 19.1725861 183.0672733
pwm am pmw QAfltliC maw
nOTrera/wt rtfttJlTO m 0.0050000 0.0050000
ROOT FILLET RADIUS MIN 0.0201191 0.0206911
WHOLE DEPTH CONSTANT 2.4000000 Lt^euvBrSfrO
CLEARANCE AT TIP OF TOOTH 0.0209483 0.0209483

BALL DIAMETER Vt iljov Bv 0.1250000

« S e f OVER-BALLS 1.9000- 1.901815.5315-15.5341
BALL (BITACT DIAMETER 1.7152- 1.716915.3622-15.3647
DIM OVER TOP-LAND 03600097 0.0742515

Figure 2.32 A 296 X 31 tooth computer output sheet. (Courtesy of American

Lohmann Corporation, Hillside, N.J.)
Gear Tooth Design

2% y 33 wpgu
DRIVEN DRIVER
PINION EXTERNAL GEAR

NUMBER OF TEETH 32J2!O0000 296.1'OOOO'i'O ’’^ANSVE" :E DIAMETRAL PITCH 19.-031611

HELIX ANGLE (DEC) 14.6704150 1^.6704150 AMSVE^BE PRESSURE ANGLE(DEG) 20.15249^0
PITCH DIAMETER 1.7O0753S 15.2552462 CENTER DISTANCE S. 4780^0

RELATIVE ROLLING SPEED (RPM) 3184Z.0000000 3549.9527027 PITCH LINE VELOCITY (FPM) 14177.3833890
MESH TORQUE (IN-LBS) 296.8956096 2663.0636495 MESH RATIO 8.96967/0
BENDING GEOMETRY FACTOR 0.5000000 0.5000000 RELATIVE HORSEPOWER PER MESH 150,0000000
BENDING STRESS (PSD 192Z0.8571549 19220.8571549 EFFECTIVE FACE WIDTH 1.6250000
BENDING LIFE (HOURS) 999999J000000 999999.0000000 STATIC TANGENTIAL LOAD (LBS) 349.1341^4
BENDING SAFETY FACTOR Z.2891799 2.2891799 DYNAMIC FACTOR 1.5316706
COMPRESSIVE STRESS (PSD 75092.5963Z51 75092.5963251 ALIGNMENT FACTOR 1.5051090
COMPRESSIVE LIFE (HOURS) 999999.0000000 999999.0000000 MODIFIED TANGENTIAL LOAD (LBS) 304,86-0-6
COMPRESSIVE SAFETY FACTOR 1.SI10973 1.8110973 SURFACE FINISH 20. !?t00000
SLIDING VELOCITY AT TIP (FPM) 1861.7184171 -3335.9694645 FLASH TEMPERATURE RISE (DEG F) 78.9349279

A.G.M.A. MATERIAL GRADE 1.0000000 1.0000000 PROFILE CONTACT RATIO 1.8454403

ALTERNATING BENDING FACTOR 1.0000000 1.0000000 FACE CONTACT RATIO 2.6274494
NUMBER OF MESHES PER REV 1.002)0000 1.0000000 MIN CONTACT LENGTH 3.0514b,0
BASE HELIX ANGLE (DEG) 13.8076960 13.8076960 MAXIMUM CONTACT LENGTH 3.1498422

OUTSIDE DIAMETER 1.777- 1.780 15.380-15.383 BACKLASH 0.0060**0

PITCH DIAMETER 1.7007538 15.2552462 CIRCULAR PITCH 0.1619114
FORM DIAMETER 1.6125344 15.1871940 LEAD ERROR (IN/IN) 0.0002^0
BASE DIAMETER 1.5966320 14.3213052 BASE PITCH 0.1519770
ROOT DIAMETER 1.523- 1.533 15.126-15.136 DEPTH TO POINT OF MAX SHEAR 0.0026780

ROLL ANGLE-MAX OUTSIDE DIA 28.2371869 22.4672316

ROLL ANGLE-ROUND EDGE DIA 27.5242396 22.3643661
ROLL ANGLE-HIGH SINGLE TOOTH 19.0158108 21.4391728
ROLL ANGLE-PITCH DIA 21.0268157 21.0268157
ROLL ANGLE-LOW SINGLE TOOTH 17.3280971 21.2510155
ROLL ANGLE-FORM DIAMETER 8.1067210 20.2229567

H P LAND THICKNESS 0.0415479 0.0347377

MAX CASE DEPTH 0.0321538 0.0261538
TRANSVERSE CIR TOOTH THICKNESS 0.0731610 0.0827504
NORMAL CIR TOOTH THICKNESS 0.0703- 0.0713 0.0796- 0.0806
NORMAL DIAMETRAL PITCH 20.0570477 20.0570477
NORMAL PRESSURE ANGLE 19.5460270 19.5460270
LEAD 20.4095271 183.0672734
ROUND EDGE RADIUS MAX 0.0050000 0.0050000
ROOT FILLET RADIUS MIN 0.0241296 0.0197955
WHOLE DEPTH CONSTANT 2.4000000 £ 4000000
CLEARANCE AT TIP OF TOOTH 0.0201918 0 .0201918

BALL DIAMETER 0.1250000 0.1250000

MEASUREMENT OVER BALLS 1.9326- 1.934615,5315-15.5341
BALL CONTACT DIAMETER 1.7558- 1.757615.3622-15.3647
DIM OVER TOP LAND 0.0762863 0.0742512

Figure 2.33 A 296 X 33 tooth computer output sheet. (Courtesy of American

Lohmann Corporation, Hillside, N.J.)
76 Gear Drive Systems

ENGINEERING DRAWING FORMAT

The engineering drawing must contain sufficient information to define the com­
ponent completely so that the manufacturing department can fabricate it and
the quality control department can inspect it. There are several elements that
should appear on the field of a gear drawing:
1. Gear blank features are usually shown in an end view and cross section, as
illustrated in Figure 2.34. It is important to specify the reference surfaces
that will locate the gear in the application. For instance, the gear in Figure
2.34 will be pressed onto a shaft; therefore, the surface C locates the gear in
the assembly and the gear tooth geometry must be accurate with respect to
this surface. Sides A and B must be parallel to each other according to the
drawing, and surface C perpendicular to A and B. The surface finish is desig­
nated by the ^sym bol. The end view shows a gear tooth and calls out which
face of the gear tooth is loaded. The X’s indicate to how many decimal
places a dimension is given.
2. A close-up view of the gear teeth as shown in Figure 2.35 defines the out­
side, pitch, form, and root diameters. It also calls out the roughness in the
root area and in the area between the form diameter and the outside diam­
eter: the active profile. The maximum undercut allowed in the root fillet

LO ADED FACE

x.xxx
T
x.xxx
T
x.xxxx
x.xxxx

ill 1B1.XXXX 111a L x X X xl

i X .X X X
* x.xxx

Figure 2.34 Gear blank dimensioning.

Gear Tooth Design 77

area is defined. If the teeth are surface-hardened by processes such as car-

burizing or nitriding, the areas to be hardened are designated as surfaces | h ]
followed by a triangle, which refers to a note that defines the case hardness
and depth. Note that in the illustration the top lands and tooth ends are not
hardened. Some designers prefer to harden these areas and therefore would
point to them in this view.
3. The tooth edges at the top land and the ends must be rounded, and Figure
2.36 illustrates how the radii of the tooth tips and edges are defined.
4. The gear material and its heat treatment must be specified and this is usually
done in a block of data on the lower right-hand side of the drawing. As an
illustration the callout for a carburized gear follows:
Material: AMS 6265 forging
Carburize surfaces fid]
Effective case depth of finished gear: 0.035 to 0.050
Case hardness: Rc 60 to 63
Core hardness: Rc 32 to 40
Per specification xxxx
Surface temper inspection per specification xxxx
Magnetic particle inspection per specification xxxx
78 Gear Drive Systems

Figure 2.36 Round edge definition.

GEAR DATA

M ET HO D OF MANUFACTURE
NO. OF T EETH •
HELSX ANGLE X X .X XX X 0
HAND OF HELSX
LEAD OF HELSX X X X .X X X X
NORM AL D8AMETRAL PSTCH X X .X X X X
NORM AL PRESS URE ANGLE x x .x x x x 0
NO RM AL CSRC. T O O T H THICK. .x x x x /.x x x x
PSTCH DIAMETER x x .x x x x
R O O T DIAMETER x x .x x x /x x .x x x
FORM DIAMETER/ MAX. x x .x x x x
WIRE OR BALL DIAMETER .x x x x
MEAS. OVE R WIRE OR BALLS x x .x x x x /x x .x x x x
AGMA QUALITY NUMBER
RUN OUT TOL. .XX X
PSTCH TOL. .x x xx
PROFILE TOL. SEE DIAGRAM
LEAD TOL. SEE DIAGRAM
MATING GEAR PART NUMBER
BACKLASH WITH MATING GEAR

Figure 2.37 Gear data block.

Gear Tooth Design 79

5. Figure 2.37 is one form of a gear data block which is applicable to both spur
and helical gears. In this format all data are given at the operating pitch
diameter. Note that the pitch diameter must be concentric to surface C
shown in Figure 2.34.

Tooth Tolerances

Note that the gear data form (Figure 2.37) has a line which calls out the AGMA
quality number. AGMA Standard 390.03 [1] specifies quality numbers identify­
ing specific tooth element tolerances. The higher the quality number, the more
precise the gearing will be and the closer the tolerances. Quality numbers range
from 3 to 15 and the standard contains a tabulation of many industrial and end
use applications and suggested quality number ranges for each. The following
table lists some sample applications:

Industry Quality number

Aerospace engines 10-13
Agriculture 3-7
Automotive 10-11
Mining 5-8
Steel 5-6

In the industries cited above and in other applications, when high-speed drives
are required or there are special considerations such as noise abatement, higher
quality numbers may be called for. It should be noted that quality classes 13,
14, and 15 are extremely difficult to achieve and prior to requiring these classes
there should be agreement between the manufacturer and user as to the method
of inspection.
The majority of critical industrial applications in fields such as the process
industries and turbomachinery will require gear units with elements that fall into
the quality number range 10 to 13. Table 2.1 presents the tolerances for these
classes. Following is a definition of each tooth tolerance element shown in
Table 2.1:
1. Runout tolerance. The variation of the pitch diameter in a direction per­
pendicular to the axis of rotation with respect to a reference surface of revo­
lution such as a bearing journal or a bore. The pitch diameter, being theo­
retical, must be indirectly measured and this can be done several ways. Two
widely used methods are:
a. Runout can be measured by indicating the position of a ball probe in
successive teeth (see Figure 2.38).
 80 Gear Drive Systems

Table 2.1 Pitch Gear Tolerances (in ten-thousandths of an inch)

RUNOUT TOLERANCE PITCH

AGMA NORMAL
QUALITY DIAMETRAL PITCH DIAMETER (INCHES) PITCH DIAMETER
NUMBER PITCH
3/4 VA 3 6 12 25 50 100 200 400 3/4 VA 3 6 12
1/2 146.5 174.5 205.8 242.7 286.3 337.6 19.0
1 88.8 104.8 124.8 147.2 173.6 204.7 241.4 14.4 16.3
2 53.9 63.5 74.9 89.2 105.2 124.1 146.3 172.6 10.9 12.3 14.0
8 4 32.7 38.5 45.4 53.6 63.8 75.2 88.7 104.6 123.4 8.3 9.3 10.6 11.9
8 19.8 23.3 27.5 32.5 38.3 45.6 53.8 63.4 74.8 88a 6.3 7.1 8.0 9.0 10.2
12 16.3 19.2 22.6 26.7 31.5 37.5 44.2 52.1 61.5 72.5 5.7 6.5 7.3 8.3 9.3
20 12.7 15.0 17.7 20.8 24.6 29.3 34.5 40.7 48.0 56.6 5.1 5.8 6.5 7.4 8.3
1/2 104.7 124.7 147.0 173.4 204.5 241.2 13.4
1 63.5 74.8 89.1 105.1 124.0 146.2 172.4 10.2 11.5
2 38.5 45.4 53.5 63.7 75.2 88.6 104.5 123.3 7.7 8.7 9.8
9 4 23.3 27.5 32.4 38.3 45.6 53.7 63.4 74.7 88.1 5.8 6.6 7.4 8.4
8 14.1 16.7 19.7 23.2 27.4 32.6 38.4 45.3 53.4 63.0 4.4 5.0 5.6 6.4 7.2
12 11.6 13.7 16.2 19.1 22.5 26.8 31.6 37.2 43.9 51.8 4.0 4.6 5.1 5.8 6.6
20 9.1 10.7 12.6 14.9 17.6 20.9 24.7 29.1 34.3 40.4 3.6 4.1 4.6 5.2 5.9
1/2 74.8 89.0 105.0 123.8 146.1 172.3 9.4
1 45.3 53.5 63.7 75.1 88.5 104.4 123.2 7.2 8.1
2 27.5 32.4 38.2 45.5 53.7 63.3 74.7 88.1 5.4 6.1 6.9
10 4 16.7 19.6 23.2 27.3 32.5 38.4 45.3 53.4 63.0 4.1 4.6 5.2 5.9
8 10.1 11.9 14.0 16.6 19.5 23.3 27.4 32.4 38.2 45.0 3.1 3.5 4.0 4.5 5.1
12 8.3 9.8 11.5 13.6 16.1 19.1 22.6 26.6 31.4 37.0 2.8 3.2 3.6 4.1 4.6
20 6.5 7.6 9.0 10.6 12.5 14.9 17.6 20.8 24.5 28.9 2.5 2.9 3.2 3.7 4.1
1/2 53.4 63.6 75.0 88.5 104.3 123.0 6.6
1 32.4 38.2 45.5 53.6 63.2 74.6 88.0 5.0 5.7
2 19.6 23.1 27.3 32.5 38.3 45.2 53.3 62.9 3.8 4.3 4.9
11 4 11.9 14.0 16.6 19.5 23.2 27.4 32.3 38.1 45.0 2.9 3.3 3.7 4.2
8 7.2 8.5 10.0 11.8 14.0 16.6 19.6 23.1 27.3 32.2 2.2 2.5 2.8 3.2 3.6
12 5.9 7.0 8.2 9.7 11.5 13.7 16.1 19.0 22.4 26.4 2.0 2.3 2.6 2.9 3.3
20 4.6 5.5 6.4 7.6 9.0 10.7 12.6 14.8 17.5 20.6 1.8 2.0 2.3 2.6 2.9
1/2 38.1 45.4 5.36 63.2 74.5 87.9 4.7
1 23.1 27.3 32.5 38.3 45.2 53.3 62.8 3.5 4.0
2 14.0 16.5 19.5 23.2 27.4 32.3 38.1 44.9 2.7 3.0 3.4
12 4 8.5 10.0 11.8 13.9 16.6 19.6 23.1 27.2 32.1 2.0 2.3 2.6 2.9
8 5.2 6.1 7.2 8.5 10.0 11.9 14.0 16.5 19.5 23.0 1.5 1.7 2.0 2.2 2.5
12 4.2 5.0 5.9 6.9 8.2 9.8 11.5 13.6 16.0 18.9 1.4 1.6 1.8 2.0 2.3
20 3.3 3.9 4.6 5.4 . 6.4 7.6 9.0 10.6 12.5 14.7 1.3 1.4 1.6 1.8 2.0

Source: AGMA Standard 390.03.

 Gear Tooth Design 81

TOLERANCE PROFILE TOLERANCE LEAD TOLERANCE

(INCHES) PITCH DIAMETER (INCHES) FACE WIDTH (INCHES)

25 50 100 200 400 3/4 VA 3 6 12 25 50 100 200 400 1and

Less 2 3 4 5
21.7 24.5 27.7 31.3 35.4 42.6 47.7 53.1 59.1 65.7 73.1
18.6 21.0 23.7 26.8 30.3 28.3 31.5 35.3 39.3 43.7 48.6 54.1
15.9 18.0 20.3 23.0 26.0 18.8 21.0 23.3 26.1 29.0 32.3 36.0 40.0
13.6 15.4 17.4 19.7 22.2 12.5 13.9 15.5 17.2 19.3 21.5 23.9 26.6 29.6 5 8 11 13 16
11.7 13.2 14.9 16.8 19.0 8.3 9.3 10.3 11.5 12.8 14.3 15.9 17.7 19.7 21.9
10.6 12.0 13.6 15.4 17.4 7.0 7.8 8.6 9.6 10.7 12.0 13.3 14.8 16.5 18.4
9.5 10.7 12.1 13.7 15.5 5.6 6.2 6.9 7.7 8.6 9.6 10.7 11.9 13.2 14.7
15.3 17.3 19.5 22.1 24.9 30.4 34.1 37.9 42.2 46.9 52.2
13.1 14.8 16.7 18.9 21.4 20.2 22.5 25.2 28.1 31.2 34.7 38.6
11.2 12.7 14.3 16.2 18.3 13.5 15.0 16.7 18.6 20.7 23.1 25.7 28.6
9.6 10.8 12.2 13.8 15.7 8.9 10.0 11.1 12.3 13.8 15.3 17.1 19.0 21.1 4 7 9 11 13
8.2 9.3 10.5 11.9 13.4 5.9 6.6 7.4 8.2 9.1 10.2 11.4 12.6 14.1 15.6
7.5 8.5 9.6 10.8 12.2 5.0 5.5 6.2 6.9 7.6 8.6 9.5 10.6 11.8 13.1
6.7 7.6 8.5 9.7 10.9 4.0 4.4 4.9 5.5 6.1 6.8 7.6 8.5 9.4 10.5
10.8 12.2 13.7 15.5 17.6 21.7 24.3 27.1 30.1 33.5 37.3
9.2 10.4 11.8 13.3 15.0 14.5 16.1 18.0 20.0 22.3 24.8 27.6
7.9 8.9 10.1 11.4 12.9 9.6 10.7 11.9 13.3 14.8 16.5 18.3 20.4
6.7 7.6 8.6 9.8 11.0 6.4 7.1 7.9 8.8 9.9 11.0 12.2 13.6 15.1 3 5 7 9 10
5.8 6.5 7.4 8.3 9.4 4.2 4.7 5.3 5.9 6.5 7.3 8.1 9.0 10.0 11.2
5.3 6.0 6.7 7.6 8.6 3.6 4.0 4.4 4.9 5.5 6.1 6.8 7.6 8.4 9.4
4.7 5.3 6.0 6.8 7.7 2.9 3.2 3.5 3.9 4.4 4.9 5.4 6.1 6.7 7.5
7.6 8.6 9.7 10.9 12.4 15.5 17.4 19.3 21.5 24.0 26.7
6.5 7.3 8.3 9.4 10.6 10.3 11.5 12.9 14.3 15.9 17.7 19.7
5.6 6.3 7.1 8.0 9.1 6.9 7.6 8.5 9.5 10.6 11.8 13.1 14.6
4.8 5.4 6.1 6.9 7.8 4.6 5.1 5.6 6.3 7.0 7.8 8.7 9.7 10.8 3 4 6 7 8
4.1 4.6 5.2 5.9 6.6 3.0 3.4 3.8 4.2 4.6 5.2 5.8 6.4 7.2 8.0
3.7 4.2 4.7 5.4 6.1 2.5 2.8 3.1 3.5 3.9 4.4 4.9 5.4 6.0 6.7
3.3 3.7 4.2 4.8 5.4 2.0 2.3 2.5 2.8 3.1 3.5 3.9 4.3 4.8 5.4
5.3 6.0 6.8 7.7 8.7 11.1 12.4 13.8 15.4 17.1 19.0
4.6 5.2 5.8 6.6 7.5 7.4 8.2 9.2 10.2 11.4 12.7 14.1
3.9 4.4 5.0 5.6 6.4 4.9 5.5 6.1 6.8 7.6 8.4 9.4 10.4
3.3 3.8 4.3 4.8 5.5 3.3 3.6 4.0 4.5 5.0 5.6 6.2 6.9 7.7 2 3 5 6 7
2.9 3.2 3.7 4.1 I 4.7 2.2 2.4 2.7 3.0 3.3 3.7 4.1 4.6 5.1 5.7
2.6 3.0 3.3 3.8 ] 4.3 1.8 2.0 2.2 2.5 2.8 3.1 3.5 3.9 4.3 4.8
2.3 2.6 3.0 3.8 1.5 1.6 1.8 2.0 2.2 2.5 2.8 3.1 3.4 3.8
"
82 Gear Drive Systems

Figure 2.38 Single probe runout check.

b. a rolling check can be conducted meshing the gear to be inspected with a

master gear of known accuracy on a fixture with a movable center dis­
tance. The variation of center distance is a measure of runout.
2. Pitch tolerance. The pitch is the theoretical distance between corresponding
points on adjacent teeth. The variation from tooth to tooth can be measured
using an instrument which employs a fixed finger and stop for consistent
positioning on successive pairs of teeth, and a movable finger which displays
pitch variations on a dial indicator or chart recorder (Figure 2.39).
3. Profile tolerance. The deviation from a true involute checked on an in­
volute profile measuring instrument. In most cases, a modified involute is
used; the drawing specification for involute modifications is discussed later
in the chapter.
4. Lead tolerance. For a spur gear the lead inspection might be considered a
check of the parallelism of the tooth with respect to the axis of rotation.
The lead of a helical gear is the axial advance of a helix for one complete
turn. Lead is checked by an instrument that advances a probe along the
tooth surface, parallel to the axis, while the gear rotates in a specified, timed
relation based on the lead.
Gear Tooth Design

PROFILE MODIFICATION

LEAD MODIFICATION

Crown

Figure 2.40 Tooth modifications.

84 Gear Drive Systems

In the data block (Figure 2.37) there are lines available for both the
AGMA quality number and the specific tolerances mentioned above. It is pos­
sible that for a specific application the designer will not choose tolerances from a
single quality number class. For instance, the designer may want to have a closer
tolerance on tooth-to-tooth spacing that on profile. In such a case the individual
tolerances can be specified on the gear data block. Even when a single quality
number is used, the tolerances can be placed on the block for reference.
When a modified involute or lead is required, a note in the gear data block
will refer to a diagram on the drawing which defines the modification. Modified
involute profiles and leads are used to attempt to compensate for deflections
during operation and tooth errors. Figure 2.40 illustrates profile (involute)
modifications and lead modification, sometimes called crowning.

Profile Modification

Tooth profiles are modified to avoid interference which can occur as the teeth
enter into or leave the mesh. The interference is a result of deflection of the gear
teeth, shafts, or gear casing due to the transmitted load or tooth discrepancies
such as spacing or profile error. For instance, if a pinion tooth is misplaced from
its theoretical position due to spacing error or because the previous tooth has
deflected under load and enters into mesh too soon, the interference with the
mating tooth will create a dynamic load which will increase tooth stresses and
system noise and vibration levels. Such interference can be eliminated by reliev­
ing the pinion and gear tooth tips or flanks or both, as shown in Figure 2.40.
Figure 2.41 illustrates how the profile modification is specified on the engineering

Figure 2.41 Modified involute diagram. The gear tooth profiles within the
tolerance bands shall not depart from a smooth and gradual convex curvature.
Gear Tooth Design 85

drawing. The diagram provides tolerance bands for the chart that will be pro­
duced when the gear is inspected on an involute measuring instrument. Figure
2.41 presents a diagram for the 31-tooth pinion described in Figure 2.32. The
amount of relief specified at the round edge diameter is 0.0003 to 0.0007. It is
impractical to specify values at the outside diameter since the round edge radius
cannot be closely controlled. The relief at the form diameter in this case is also
0.0003 to 0.0007. The tip and flank reliefs commence at the highest and lowest
points of single tooth contact.
It should be recognized that the benefits from profile modification can be
realized only if the teeth are accurately manufactured to tolerances less than the
modifications specified. It would be pointless to have a 0.0005 tip modification
if the tooth spacing were allowed to vary by 0.001. An estimate of the amount
of modification required for a given application may be found in Ref. 2.
The modification at the first point of contact is given as:
3 5 X 1 0 -7
Modification = driving load (lb) X —:---------------
face width (in.)
To achieve this modification, material must be removed from the tip of the
driven gear or the flank of the driving gear or both. If material is removed from
both, the total modification is split between the two meshing teeth.
The modification at the last point of contact is given as:
9 0 X 1 O-7
Modification = driving load (lb) X :-------------
face width (in.)
To achieve this modification, material must be removed from the tip of the
driving gear or the flank of the driven gear or both.
The foregoing estimates for profile modification are offered as starting
points for a design. The final tooth modifications are arrived at through develop­
ment by observing operating results.
Lightly loaded gears may not require profile modification. If a simple
involute tolerance is called out on the gear data block, Figure 2.42 shows how
this tolerance is to be interpreted. Assume a profile tolerance of 0.0008 in. The
measured profile must fall within the checked area of the diagram in Figure
2.42. A true involute would be a straight line on the diagram. If the involute is
plus, the line on the diagram will slant up toward the left. A minus involute is
depicted by a dashed line on the diagram. In general, a plus involute tends
toward an interference condition; therefore, the minus involute is more desirable.

Lead (Axial) Modifications

In theory, when gear teeth mesh the faces will be parallel to each other and the
load will be distributed across the full face width. In practice, however, a
86 Gear Drive Systems

Figure 2.43 Lead diagram.

Gear Tooth Design 87

tolerance must be given. For spur gears it will be a tolerance on the parallelism
of the tooth with the axis of rotation. For helical gears it is called a lead
tolerance. Figure 2.43 shows how this tolerance is interpreted. The solid line is
the theoretical trace and the dashed line is the measurement as recorded by a
lead checking instrument. The diagram shows a 0.001-in. variation. If this is a
right-hand helical gear, because the variation is off to the right, the measured
helix angle is greater than the theoretical value.
Tooth faces tend not to be exactly parallel in operation not only because
of tooth errors but also due to deflections of shafts, bearings, and casings. The
load, therefore, may be concentrated on an end rather than distributed evenly
across the face. To alleviate end loading, a lead modification, sometimes called a
crown, is used. As shown in Figure 2.40, the crown relieves the tooth ends and
avoids a heavy concentration of load in these areas. Figure 2.44 illustrates how a
crowned tooth is specified on the engineering drawing. In this example the relief
at either end of the face width is 0.0004 to 0.0008, blending smoothly into the
flat at the center of the tooth. The amount of crowning generally is on the order
of 0.001 in., but like the profile modification must be finally developed by
observation of test results. In some cases deflections are such that only one end
of the tooth need be crowned. On occasion two mating gears are designed with
slightly different helix angles which become parallel as the system deflects
under load.

Figure 2.44 Crowned lead diagram. The lead contour within the tolerance
bands shall not depart from a sm ooth and gradual convex curvature.
88 Gear Drive Systems

SPLINE DESIGN

Splines are used in mechanical systems to transmit torque and motion from one
shaft to another. A spline connection consists of a set of external gear teeth
arranged in a circle which fit into a corresponding set of internal gear teeth.
Splines provide a strong, compact method of connection which can accommo­
date some misalignment.
In general use today are involute splines of 30° pressure angle with stub
teeth. Stub teeth have short addendums of l/2(diametral pitch) rather than the
conventional 1/diametral pitch. Because of this the spline pitch is conventionally
given as a fraction (e.g., 12/24, the numerator being the diametral pitch, which is
the number of spline teeth per inch of pitch diameter and which controls the
circular pitch and basic space width or tooth thickness. The denominator is
known as the stub pitch and is always twice the numerator. The tooth adden­
dum is 1/stub pitch.
Spline teeth are of the involute form because of tooling advantages. A
single hob or shaper cutter can generate all numbers of teeth of a given diametral
pitch. Stub teeth and 30° pressure angles are used for ease of machining. The
relatively high pressure angle increases tool life because the tool has more clear­
ance behind the cutting edge. Also, higher cutting speeds are possible. The stub
tooth is advantageous for broaching internal splines and for rolling of teeth.
Generally, splines are designed to ANSI Standard B92.1a [3]. In addition
to 30° pressure angles the standard presents dimensioning systems for 37.5° and
45° pressure angle splines, which are sometimes known as serrations. Figure 2.45
taken from the standard shows spline tooth nomenclature and how spline data
are presented on the engineering drawing.
Two root fillet configurations are possible, as shown in Figure 2.45. One is
the flat root spline, in which fillets join the arcs of major or minor circles to the
tooth sides. The other is the full fillet root spline, in which a single fillet in the
general form of an arc joins the sides of adjacent teeth. The full fillet root form
is stronger and should be used if appreciable torque is transmitted through the
spline.
There are two types of fits possible with mating splines. One is a side fit
where the mating members contact on the sides of the teeth only and there is
clearance at the major diameters. When using a side fit spline if more accurate
centralization of the shafts is desired, this can be accomplished by the use of
shaft shoulders, as shown in Figure 2.46. It is also possible to have a major
diameter fit where the mating members contact at their major diameters and the
tooth sides act only as drivers. In this case the standard provides for increased
clearance at the sides to ensure that all centering will be at the major diameters.
To be sure that two mating splines will fit together with minimum clear­
ance, the concept of effective and actual tooth space and tooth thickness
 Gear Tooth Design 89

INTERNAL SPLINE

EXTERNAL SPLINE

(CIRCULAR)

DRAWING DATA

INTERNAL INVOLUTE SPLINE DATA EXTERNAL INVOLUTE SPLINE DATA

-ILLET ROOT SIDE FIT FILLET ROOT SIDE FIT

NUMBER OF TEETH XX NUMBER OF TEETH XX

SPLINE PITCH XX/ XX SPLINE PITCH XX/ XX

o
o

PRESSURE ANGLE PRESSURE ANGLE

o
CO
o

CO
BASE DIAMETER X . X X X X X X REF. BASE DIAMETER X. X X X X X X REF.
PITCH DIAMETER X .-X X X X X X REF. PITCH DIAMETER X . X XX X XX REF.
MAJOR DIAMETER X. XXX MAX. MAJOR DIAMETER X. XXX/ X. XXX
FORM DIAMETER X. XXX FORM DIAMETER X. XXX
MINOR DIAMETER X. XXX/ X. XXX MINOR DIAMETER X. XXX MIN.

CIRCULAR SPACE WIDTH CIRCULAR TOOTH THICKNESS

MAX ACTUAL X . XXXX MAX EFFECTIVE X. XXXX
MIN EFFECTIVE X. XXXX MIN ACTUAL X. XXXX

MAX MEAS.BETW. PINS X . XXXX REF. MIN MEAS. OVER PINS X. XXXX REF.
PIN DIAMETER X . XXXX PIN DIAMETER X. XXXX

Figure 2.45 Spline nomenclature and drawing data.

dimensions is used in spline tooth dimension systems. To understand this con­

cept, imagine an internal spline with each tooth space width exactly half a
circular pitch and the mating external spline with each tooth thickness exactly
half a circular pitch. It would seem that these splines would fit perfectly; how­
ever, because of such tooth errors as spacing, profile, out of round, and lead, the
pair probably cannot be assembled. Because of these errors the spline teeth will
not be in their theoretical locations on the pitch circle and at some point or
points there will be interference between the internal and external teeth. To
overcome this problem, all space widths of the internal spline must be widened
by the amount of interference caused by tooth errors and all tooth thicknesses
90 Gear Drive Systems

of the external spline must be thinned. This concept leads to four dimensions for
space width and tooth thickness:
Minimum effective space width = V2 circular pitch (0 = 30°)
Maximum effective space width
Minimum actual space width
Maximum actual space width
Maximum effective tooth thickness = Vi circular pitch (0 = 30°)
Minimum effective tooth thickness
Maximum actual tooth thickness
Minimum actual tooth thickness
The spline teeth are machined to the actual space width or tooth thickness
dimensions which can be checked by the use of gages or measurements over pins.
The effective dimensions are checked by gages. There are four machining toler­
ance classes set up for the effective and actual space widths and tooth thick­
nesses which result in varying degrees of clearance.
It must be remembered that the ability to assemble the spline is not the
only criterion in critical applications. When significant loads are transmitted, or at
high speed, the tooth geometry of the splined connection may have to be closely
controlled and elements such as profile, lead, and surface texture specified.
Gear Tooth Design 91

In some cases where tooth bearing surface is important it may be desirable to use
full-depth teeth. Full-depth splines would not use a 30° pressure angle since the
teeth would be too pointed but would have a conventional 20° pressure angle
or less.
There are applications where it is desirable to transmit thrust through a
splined connection. For instance, a turbine wheel may be connected to a
helical gear and the thrust of the wheel offset by the gear thrust. In such a case
a helical tooth spline is effective. Figure 2.47 illustrates such a system. The
spline helix angle is chosen such that the spline thrust exceeds the gear thrust
and the shaft system locks up with the turbine shaft, bottoming out in the gear
shaft shoulder. The net thrust in the system is then the turbine thrust minus
the gear thrust, which is reacted by the ball bearing. The bearing loading, there­
fore, is greatly reduced from the case where the thrusts are not offset.

REFERENCES

1. AGMA Gear Handbook 390.03, Vol. 1, Gear Classification, Materials and

Measuring Methods for Unassembled Gears, American Gear Manufacturers
Association, Arlington, Va., January 1973.
2. Dudley, D. W., Gear Handbook, McGraw-Hill, New York, 1962, pp. 5-23.
3. ANSI Standard B92.1a, Involute Splines and Inspection, Society of Auto­
motive Engineers, Warrendale, Pa., 1976.
3
GEARBOX RATING

The rating of a gearbox is determined by the loads the gearbox components are
capable of transmitting. In some cases where a system is operating continuously
at a uniform load such as an electric motor driving &fan, the loading is simple to
predict and component analysis can be based on a continuous horsepower trans­
mitted at steady speed. Some applications, however, experience variable loading
such as high starting torque or shock loads and these conditions must be con­
sidered in the gearbox design.
Organizations such as the American Gear Manufacturers Association
(AGMA) and the American Petroleum Institute (API) issue Standards that define
gear rating procedures. AGMA Standard 420.04 [1] covers enclosed drives with
pitch line velocities not exceeding 5000 fpm or pinion speeds not exceeding
3600 rpm. Higher-speed enclosed drives are covered by AGMA Standard 421.06
[2]. The general AGMA Standard for gear rating is 218.01 [4].
The rating methods used in these standards are discussed in this chapter.
Before going into detail, an overview of the procedure one would use in rating a
gearbox follows:
1. Gear tooth rating. The first step in determining a gearbox rating is to evalu­
ate the tooth meshes. The classical gear tooth limitations that are calculated
are the fatigue phenomena of breakage and pitting. Tooth breakage is anal­
yzed by calculating the bending stress in the root fillet area and comparing it
against a material strength rating. Pitting is analyzed by calculating the com­
pressive stress at the tooth contact and comparing it against a material dura­
bility rating. A third gear tooth limitation encountered in high-speed gearing
is instability of the lubricant film, allowing metal-to-metal contact leading to

93
94 Gear Drive Systems

scoring. The failure modes of tooth breakage, pitting, and scoring are de­
scribed in Chapter 12. Their analysis is covered later in this chapter.
2. Bearing rating. Bearing ratings may be the limiting factor in determining the
load a gearbox can transmit. A decision must be made as to the minimum
acceptable L i0 life desired for antifriction bearings or the maximum loading
acceptable for journal bearings. The analysis and rating of bearings is pre­
sented in Chapter 4.
3. Thermal rating. Gear stresses or bearing lives usually determine the mechan­
ical rating of a gearbox. In addition to the mechanical rating, gearboxes
which do not use external cooling have a thermal capacity. This is defined in
AGMA Standard 420.04 as the horsepower a unit will transmit continu­
ously for 3 hr or more without exceeding a sump temperature of 200° F or a
sump temperature rise of 100°F over ambient. If these thermal limits are
exceeded, external cooling must be provided. Gearbox thermal ratings and
lubrication systems are discussed in Chapter 5.
4. Shaft rating. Consideration must be given to gearbox components other
than gears and bearings. Shafting, key ways, splines, and so on, must be
analyzed to assure satisfactory performance under load. These machine
elements are discussed later in the chapter.
The four points above are the obvious design details that must be addressed;
however, there are many other details, such as housing and shaft deflections,
critical speeds, and thermal expansion, of which the experienced gearbox de­
signer is aware. Generally, in the procurement of a gearbox, gear ratings, bearing
lives, and lubrication details are documented and the finer points of gearbox
design are left to the manufacturer.

TOOTH LOADS

To calculate gear and shaft stresses and bearing lives, the gear tooth loads must
be developed. Figure 3.1 illustrates the load diagram on a spur gear tooth. The
gear torque is
__ 63,025 hp
n rpm
where
Tq = torque, in.-lb
hp = horsepower
rpm = gear rotational speed
The total transmitted tooth load R acts normal to the involute profile. The com­
ponent of R transmitting the torque is the tangential load:
Gearbox Rating 95

where
WT = tangential load, lb
PD = gear pitch diameter, in.
The total transmitted load is

cos 0T
where
R = total transmitted (resultant) force, lb
0T = transverse pressure angle, deg
As shown in Figure 3.1, the force R is resolved into the tangential load and a
separating load:
S = WT tan 0
where S is the separating load, in pounds. In the case of helical gears, the resul­
tant force R is in the normal plane. To resolve R into a tangential and separating
force, the geometry shown in Figure 3.2 is used. There is also a thrust force
generated since the resultant force is at an angle 0 to the tangential force:
96 Gear Drive Systems

WT = R cos 0N cos 0
S = R sin 0N = WT tan 0T
T = R cos 0N sin 0 = WT tan 0
where
0 = helix angle, deg
0N = normal pressure angle, deg
T = thrust load, lb

STRENGTH RATING

The strength rating of a gear tooth concerns itself with the bending stress (Figure
3.3) in the tooth fillet, where fatigue cracks initiate and propogate resulting in

TOTAL
TRANSMITTED
LOAD

BENDING S T R E S S

Figure 3.3 Bending stress criterion for strength rating.

 Gearbox Rating 97

fracture of teeth or portions of teeth. The fundamental equation for bending

stress in a gear tooth is [4]

where
St = tensile or bending stress, psi
WT = transmitted tangential load, lb
Pd = transverse diametral pitch, in.-1
J = geometry factor
The geometry factor J is an index of the following:
Tooth geometry in the root fillet area
Stress concentration in the root fillet
Load sharing between teeth
The position at which the most damaging load is applied
GEOMETRY FACTOR J

Figure 3.4a Geometry factors, 20° spur gears, standard addendum (From Ref. 5.)
 98 Gear Drive Systems
GENERATING RACK

nNs I ! 5 T
VALUE FOR Z IS FOR AN ELEMENT OF INDICATED
NUMBERS OF TEETH 8 A 75 TOOTH MATE.
NORMAL TOOTH THICKNESS OF PINION AND GEAR
TOOTH EACH REDUCED .024" TO PROVIDE .04 8"
TOTAL BACKLASH FOR I Pnd
FACTORJ

FACTORS ARE FOR

GROUND TEETH
CUT WITH A PRE­
GRIND HOB.
GEOMETRY

10° 15° 20°

HELIX ANGLE ^

Figure 3.4b Geometry factors, 20° helical gears, standard addendum. (From
Ref. 5.)

THE MODIFYING FACTOR CAN BE APPLIED TO THE

J - FACTOR WHEN OTHER THAN 75 TEETH ARE USED
IN THE MATING ELEMENT.

1.05

a:
eg 1.00
r
w
-
500
150
75
50
p yj
o sm 30
yj
if 20
.95 €3
o eg Z
z — !
yj h-
1 [ - m <
! 1" T I j 2 2
o .90 i i 3
o i |
~ |
Z
1
i . ...
-
_
' ‘"1 "T " —
.. i _

.85
0® 5®
-

10®
!

115® 20®
-

25® 30®
- - - -

35®
HELIX ANGLE

Figure 3.4c Modifying factor for helical gear geometry factors. (From Ref. 5.)
Gearbox Rating 99

Geometry factors may be arrived at by graphical layout of the tooth form

or computer analysis of the graphical procedure. Reference 5 presents a method
for calculating the geometry factor. Figures 3.4a, 3.4b, and 3.4c present geometry
factors for 20° pressure angle spur and helical gears. The geometry factor is
strongly dependent on the cutting tool geometry and these figures are for
hobbed gears. In general, spur gear geometry factors vary from approximately
0.35 to 0.45 and helical gear geometry factors have values from approximately
0.4 to 0.6.

BENDING STRESS RATING

In the AGMA rating system the basic bending stress is modified by several
factors that deal with the characteristics of a specific application:

(3.1)
Kv F I
where
Ka = application factor. This factor takes into account the roughness or
smoothness of the driving and driven equipment. When no overloads
are anticipated Ka may be taken as 1.0. For very rough operation Ka
may be 2.25 or greater.
Kv = dynamic factor. The dynamic factor represents the ratio between the
maximum dynamic load on the gear teeth and the static calculated
load. Gear teeth generate dynamic loads due to component
geometry errors which result in gear accelerations and decelerations.
Although the dynamic factor is used as a multiplier in the stress
equation, the dynamic load is actually an incremental force which
adds to the tangential force.
The dynamic factor increases with increasing pitch line velocity and decreases
with increasing tooth accuracy and increased tooth loading. As the tooth loading
is increased, the tooth deflections tend to overshadow tooth geometry errors and
the dynamic load is a smaller percentage of the total load. Figure 3.5 illustrates
this trend. The data shown were developed from a test conducted on a helical
planetary gearset transmitting 1100 hp at 21,000 rpm-in. The sun gear was strain
gaged in the root to measure tooth loading. Gear quality was AGMA Quality
Class 12 [6]. For gears of lower quality classes operating at lower speeds, the
following estimates can be used for dynamic factors [4]:
100 Gear Drive Systems

PITCH LINE VEL O CIT Y ( fpm )

Figure 3.5 Dynamic factors for accurate gearing.

where PLV is the pitch line velocity in fpm, for AGMA Quality Class 11 gearing
operating at pitch line velocities less than 8000 fpm with rigid accurate
mountings;

Ky = /
v \84 + PLV
for AGMA Quality Class 10 gearing operating at pitch line velocities of less than
6000 fpm and
„ _ , 70.7 \0.63
JvT. —
v \70.7 + PLV
for AGMA Quality Class 8 gearing operating at pitch line velocities of less than
5000 fpm.
There are analytical methods for calculating dynamic loads [7]. The tooth
stiffness and mass are determined, and assuming the magnitude of tooth errors,
an estimate of the dynamic load can be arrived at.
K s = size factor. The size factor reflects nonuniformity of material prop­
erties which become more prevalent as the size of a gear increases;
however, standard size factors have not yet been established and Ks
is usually taken as 1.0.
Gearbox Rating 101

Table 3.1 Allowable Bending Stress Sat for Gear Steels

Material hardness Sat(PsO
180 Bhn 25,000-33,000
300 Bhn 36,000-47,000
400 Bhn 42,000-56,000
Carburized Rc 55 55,000-65,000
Carburized Rc 60 55,000-70,000
Nitrided Rc 60 38,000-48,000
Source: Ref. 4.

Km = load distribution factor. The load distribution factor accounts for

inaccuracies in the bearing bore locations leading to misalign­
ment of the axes of rotation, alignment errors due to gear tooth
inaccuracies, and deflections due to load or thermal distortion.
For face widths less than 2.0 in., accurate gears and mountings
and stiff housings a Km as low as 1.1 may be used. In cases where
poor alignment is anticipated Km may equal 2.0 or more.
The relation of calculated bending stress to the allowable stress of the
material is [4]

s < ^
1 — KtKr
where
Sat = allowable material stress, psi (see Table 3.1)
Kj = life factor

Table 3.2 Life Factor-Kj

Number of cycles 160 Bhn 250 Bhn 400 Bhn Case carb.
Up to 1000 1.6 2.4 3.4 2.7
10,000 1.4 1.9 2.4 2.1
100,000 1.25 1.5 1.7 1.6
1 million 1.1 1.1 1.2 1.2
10 million 1.0 1.0 1.0 1.0

Source: Ref. 4.
102 Gear Drive Systems

The life factor adjusts the allowable stress for the required number of
operation. Table 3.2 defines values of Kp Beyond IQ7 cycles the allowable stress
may have to be further derated. A value of Kj of 0.8 might be chosen for 1010
cycles.

Kt = temperature factor. The temperature factor is usually taken as 1.0

unless the oil or gear blank temperature exceeds 250°F.
Kr = reliability factor. The allowable bending stress values in Table 3.1
reflect a failure probability of fewer than 1 in 100 in 107 load
cycles. If a lower statistical probability of failure is desired, Kr must
be greater than 1.0. A Kr of 1.25 may reflect a failure probability of
1 in a 1000 and a Kr of 1.5 may reflect a failure probability of 1 in
10000.

Power Rating

Quite often, gears are rated on the basis of power. Equation (3.1) can be manip-
ulated to the following:

F J SatKj
_ ----- _JLJL (3.2)
126,000Ka Km KsPd KrKt
where

Pat = allowable transmitted power on the basis of bending strength, hp

np = pinion speed (high-speed member), rpm
PD = pinion pitch diameter, in.

Overloads

When a gear is subjected to infrequent momentary high overloads, the transient

stress should be compared to the yield strength of the material. A factor of
safety Kr, ranging from 1.33 for conventional industrial gearing to 3.0 for very
high reliability, should be applied. Examples of infrequent overloads are equip­
ment that experiences high starting loads once a day or construction machinery
that occasionally stalls, incurring high power drain at low speed.

Reverse Bending

When gear teeth experience reverse bending (loading in both directions), such as
in an idler or planet gear, 70% of the allowable fatigue strength should be used.
Gearbox Rating 103

DURABILITY RATING

The durability rating of gear teeth concerns itself with fatigue pitting resistance.
Equations based on work by Herz are used to calculate contact (compressive)
stresses between the mating gear teeth (Figure 3.6). Gear tooth contact condi­
tions are similar to those between two cylinders except that on gear teeth the
radii of curvature are continuously changing. A specific mesh point, such as at
the pitch diameters, is chosen at which to make the calculation. Although it is
the surface stress which is calculated, subsurface shear stresses proportional to
the surface compressive stress are the actual cause of crack initiation. The funda­
mental equation for compressive stress in a gear mesh is [4]

where
Ep = pinion modulus of elasticity, psi
Eg = gear modulus of elasticity, psi
/ip = pinion Poisson’s ratio
ixg = gear Poisson’s ratio
For steel-on-steel gears the elastic coefficient Cp = 2300 psi.
I = geometry factor
The geometry factor deals with the radii of curvature at the point of contact and
the load sharing between teeth:

Figure 3.6 Compressive stress criterion for durability rating.

104 Gear Drive Systems

Cc
I = _L
M„

where
Cc = curvature factor
Mn = load sharing ratio
For helical gears the compressive stress is usually calculated at the pitch line,
where
cos 0t sin 0t / Me \
C = ---------------- I 2— | + for external gear mesh, - for internal
2.0 \Mg ± 1 . 0 / gearmesh

where
0t = transverse pressure angle, deg
Mg = gear ratio (number of gear teeth divided by number of pinion teeth,
always greater than 1.0)
For spur gears the compressive stress is usually calculated at the start of single
tooth contact. The curvature factor Cc must be multiplied by a modification
factor Cv :

Ri R 2
Cx =
Rp Rg
where
Ri = radius of curvature at the lowest point of single tooth contact on
the pinion, in.
R2 = radius of curvature at the highest point of single tooth contact on
the gear, in.
Rp = radius of curvature at the pinion pitch diameter, in.
R q = radius of curvature at the gear pitch diameter, in.
The method of calculation of these radii of curvature was discussed in Chapter 2.
Mn =load-sharing ratio, which depends on theprofile and face contact
ratios
Mn = 1 .0 for spur gears
Mn = F/L j^ h for helical gears
Lmin = total length of lines of contact(minimum), in.
A procedure for calculating Lmin was presented in Chapter 2. An estimate for
Lmin which holds for most helical gears when the face contact ratio exceeds 2.0,
or when the face contact ratio or transverse contact ratio is an integer of 1.0 or
greater, is
Gearbox Rating 105

0.95Z
where
pn = normal base pitch, in.
Z = length of action in the transverse plane, in. (see Figure 2.15)

COMPRESSIVE STRESS RATING

In the AGMA rating system the basic compressive stress is modified by several
factors that deal with characteristics of a specific application:

(3.3)

where Cf is the surface condition factor. This factor depends on the tooth finish,
residual stresses, and work-hardening effects. For good-quality gearing it is taken
as 1.0. The Ca, Cv, Cs, and Cm factors have the same values as the corresponding
Ka, Kv, Ks, and Km factors discussed previously.
The relation of calculated compressive stress to the allowable stress of the
material is

where
Sac = allowable contact stress, psi (see Table 3.3)
Ch = hardness ratio factor. When the pinion is significantly harder than
the gear, there is a work-hardening effect on the gear and the gear

Table 3.3 Allowable Contact Stress Sac for Gear Steels

Material hardness Sac (Psi)
180 Bhn 85,000-95,000
300 Bhn 120.000-135,000
400 Bhn 155.000-170,000
Carburized Rc 55 180.000-200,000
Carburized Rc 60 200.000-225,000
Nitrided Rc 60 192.000-216,000
Source: Ref. 4.
106 Gear Drive Systems

BRINE LL OF PINION
K * WHEN K < 1.2 USE G ^ l .0 0
BRINE LL OF GEAR

1.14
fill ;![| '1‘
IIS1 Sf
irrrt •+141IldH
i:f 1H;
T:rtr iftf m -.-. : rf: K *1.7
i t f StT~b b
1.12- 7rP-44-L IM i.i+4 *+—
ul; tp b - b l S K *1.6
ij|[
:l;!r
...+:;
I .IO« m m K sf .5
Iljl ifij ti!j l i i f e
.....1-
:::1 .!!,11It1 f T™ l l l l
1,:: Hfi ti3-vA | $ l i K *1 .4
flri ijil ,,Jf/ .. r ,
FACTOR

1.08
liiL s k i
■J , ;if-tff
m M
rti1—l l l i
rriTori K *1.3
1.06- tr i'f .jr .
RATIO

Ttfr-rt]ffU
trf.it: -t"th
l4+-l'fT'*tTt
\'yjt
trtt-IItsgjiHrffmi K s 1.2
HARDNESS

1.04- ::trf
'::::■■;t:; jij p pTIff--H-tHti-:
■■

■-j pt 1
it f1!rpJiMjlff:
IT
TTntI’ftt:
ItIIr"i-K
tM fm
-t-t-'ttrt
-f:h
- lb ;H { 4l|riffii
Itfr
1. 0 2 - rr: rtt{ 1■ l bttt rffifrrttfrfe
b ill :tf4f !tr r^rr-rr; rrt:.:r;tt
rrbrttft 1 bibb Jtt'i rf.li ;::: I:.;:
- w :;;; Lul bibb pfjjjj. iifHfii iirrFut
jjtf :;jt
1.00-
12 16 20
SINGLE REDUCTION GEAR RATIO

Figure 3.7 Hardness ratio factor. (From Ref. 4.)

durability rating is increased. This effect increases with the reduc­

tion ratio as shown in Figure 3.7.
Ci = life factor. The life factor adjusts the allowable stress for the re­
quired number of cycles of operation. A factor of 1.0 is used for
107 cycles. For 104 cycles and less a factor of 1.5 can be used.
Insufficient data are available to define Cx beyond 107 cycles;
however, it appears there is no well-defined endurance limit and
allowable stress should be decreased as load cycles accumulate over
107. A life factor of 0.7 might be used for 1010 cycles.
Gearbox Rating 107

The temperature factor Ct and the reliability factor Cr have the same values as
the corresponding Kt and Kr factors discussed previously.

Power Rating

Quite often, gears are rated on the basis of power. Equation (3.3) can be manip­
ulated to the following:
n„F ICV /S ac • PD C,Ch \ 2
Pac = ---- 2-------------- (— ------- \ hp (3.4
126,000 CsCmCfCa \ Cp CtCr /
where Pac is the allowable transmitted horsepower on the basis of compressive
stress.

AGMA STANDARDS FOR ENCLOSED DRIVE RATINGS

AGMA Standard 420.04 [1] covers enclosed drives with pitch line velocities not
exceeding 5000 fpm and pinion speeds not exceeding 3600 rpm. Higher-speed
enclosed drives are covered by AGMA Standard 421.06 [2], which deals with
helical and herringbone gear units. These Standards basically use Eqs. (3.2) and
(3.4) for strength and durability rating but simplify the many modifying factors
and arrive at a service factor based on transmitted power rather than a reli­
ability factor based on stress.
For the strength rating Standard 420.04 states that
K ,K 2 K3J
Pat
Pd

K = nP ’ PD ‘ Kv
1 126,000

K2 = —
Km
K3 = SatK.
where
Pat = horsepower rating based on tooth strength
J = geometry factor (Figures 3.4a, 3.4b, and 3.4c)
Pd = diametral pitch—transverse, in.-1
np = pinion rpm
PD = pinion pitch diameter, in.
Kv = 50/(50 + \/PLV) for spur gears
108 Gear Drive Systems

Figure 3.8 K2 factor for strength rating of spur and helical gearings. (From
Ref. 1.)

Kv = \ / 78/(7 8 + v/pLV) for helical gears

PLV = pitch line velocity, fpm
K2 = face width alignment factor as specified in Figure 3.8
K3 = allowable fatigue stress times life factor as specified in Figures 3.9
and 3.10
For the durability rating Standard 420.04 states:

^ac = C1C2C3C4 (3.5)

C = nP 8 P° 2 9Cv
1 126,000

c3 = I(Sac/Cp)2

C4 = (C ,)2
Gearbox Rating 109

BRINELL HARDNESS OF CYCLES

NUMBER
REQUIRED

Figure 3.9 K3 factor for strength rating of spur gears. (From Ref. 1.)

where
Pac = horsepower rating based on tooth durability
np = pinion rpm
PD = pinion pitch diameter, in.
Cv = 50/(50+ \/PLV) for spur gears
Cv = 78/(78 + VPLV) for helical gears
PLV = pitch line velocity, fpm
C2 = face width/alignment factor as specified in Figure 3.11
I = geometry factor (see also “Durability” section)
(Sac/Cp)2 = allowable contact stress/elastic coefficient as specified in
Figure 3.12
C4 = life factor2 as defined in Figure 3.13
110 Gear Drive Systems

c y c le s
STEEL

of
number
FOR

req'd

BRINELL HARDNESS

Figure 3.10 K 3 factor for strength rating of helical gears. (From Ref. 1.)
 u 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
NET FACE WIDTH (IN.)
Figure 3.11 C2 factor for durability rating. (From Ref. 1.)

I ll
 175 200 225 250 275 300 325 350 375 400 425 450 475

MINIMUM GEAR HARDNESS (Bhn)

VALUES ARE TO BE TAKEN FROM THE ABOVE CURVE FOR THE MINIMUM HARDNESS SPECIFIED FOR THE GEAR.
VALUES FOR SUGGESTED GEAR AND PINION HARDNESS COMBINATIONS ARE TABULATED BELOW FOR CONVENIENCE

MINIMUM BRINELL HARDNESS

GEAR 180 210 225 24 5 2 55 270 285 300 335 35 0 375 55 Rc 58 Rc
PINION 210 245 265 285 295 310 325 340 375 390 415 55RC 58 Rc

1750 2100 2300 2560 2700 2950 3200 3460 4200 4460 4820 6200 6600
(*)■
Figure 3.12 C3 factor for durability rating. (From Ref. 1.)
Gearbox Rating

Figure 3.13 C4 factor for durability rating. (From Ref. 1.)

HIGH-SPEED GEARING

Strength

The high-speed standard, Standard 421.06, uses Eqs. (3.2) as it stands except
that the dynamic factor Kv is defined as follows:
Kv = ^78/(78 + v/PLV) for pitch line velocities below 7000 fpm
Kv = 0.695 for pitch line velocities 7000 fpm and higher

Durability

The 421.06 Standard uses Eq. (3.5):

Pac = Ci C2 C 3 C4

with three modifications:

1. C4 = 1.0.
2. The dynamic factor Cv is 0.48 for pitch line velocities of 7000 fpm and
higher.
3. The C2 factor is defined in Figure 3.14.

Service Factors

When PQ di+ and Par

d^ are calculated, the rating of the gearbox will be based on the
smaller of the strength and durability ratings. A service factor must now be
11 ;

10

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
NET FACE W ID T H (IN.)

C2 factor for high-speed-unit durability rating. (From Ref. 2.)

Gearbox Rating 115

Table 3.4 Service Factors for High-Speed Units

Service factor
Prime mover
Internal
combustion
engine
Application Motor Turbine (multicylinder)
Blowers
Centrifugal 1.4 1.6 1.7
Lobe 1.7 1.7 2.0
Compressors
Centrifugal—process gas except air
conditioning 1.3 1.5 1.6
Centrifugal—air conditioning service 1.2 1.4 1.5
Centrifugal—air or pipeline service 1.4 1.6 1.7
Rotary—axial flow, all types 1.4 1.6 1.7
Rotary—liquid piston (Nash) 1.7 1.7 2.0
Rotary—lobe-radial flow 1.7 1.7 2.0
Reciprocating—3 or more cylinders 1.7 1.7 2.0
Reciprocating—2 cylinders 2.0 2.0 2.3
Dynamometer—test stand 1.1 1.1 1.3
Fans
Centrifugal 1.4 1.6 1.7
Forced draft 1.4 1.6 1.7
Induced draft 1.7 2.0 2.2
Industrial and mine (large with
frequent start cycles) 1.7 2.0 2.2
Generators and Exciters
Base load or continuous 1.1 1.1 1.3
Peak duty cycle 1.3 1.3 1.7
Pumps
Centrifugal (all service except as
listed below) 1.3 1.5 1.7
Centrifugal—boiler feed 1.7 2.0
Centrifugal—descaling (with
surge tank) 2.0 2.0
Centrifugal—hot oil 1.5 1.7
Centrifugal—pipeline 1.5 1.7 2.0
Centrifugal—water works 1.5 1.7 2.0
Dredge 2.0 2.4 2.5
Rotary—axial flow, all types 1.5 1.5 1.8
Rotary—gear 1.5 1.5 1.8
Rotary—liquid piston 1.7 1.7 2.0
Rotary—lobe 1.7 1.7 2.0
116 Gear Drive Systems

Table 3.4 (Continued)

Service factor
Prime mover
Internal
combustion
engine
Application Motor Turbine (multicylinder)
Rotary—lobe 1.7 1.7 2.0
Rotary—sliding vane 1.5 1.5 1.8
Reciprocating—3 cylinders or more 1.7 1.7 2.0
Reciprocating—2 cylinders 2.0 2.0 2.3
Paper industry
Jordan or refiner 1.5 1.5
Paper machine—line shaft 1.3 1.3
Paper machine—sectional drive 1.5
Pulp beater 1.5
Sugar industry
Cane knife 1.5 1.5 1.8
Centrifugal 1.5 1.7 2.0
MiU 1.7 1.7 2.0
Source: Ref. 2.

assigned to the application in order to determine the horsepower that the unit
will be rated against. The specified or actual horsepower transmitted through the
gearbox is multiplied by the service factor and called the equivalent or service
horsepower. The smaller of Pac and Pat must equal or exceed the equivalent
horsepower:
smaller of P„r or Pn+
Service factor = ac at
specified horsepower
The AGMA and API Standards [1-3] contain tables of recommended
service factors for various applications. Table 3.4 illustrates the numerical
values of service factors for high-speed units. The API Standard has very
similar values. Lower-speed gearboxes covered by AGMA Standard 420.04 have
service factors as shown in Table 3.5, which takes into account the duration of
service, prime mover characteristics, and driven machine characteristics.
The AGMA and API standards are good basic guidelines which, when fol­
lowed, result in gearboxes that perform successfully. They provide a baseline
with which to judge competing designs. The enclosed drive standards [1-3] will
give more conservative results than the general standard [4], but this is to be
Gearbox Rating 117

Table 3.5 Service Factors for Low-Speed Units

Driven machine
load classification
Moderate Heavy
Prime mover Duration of service Uniform shock shock
Electric motor, Occasional V2 hr per day 0.50 0.80 1.25
steam turbine, Intermittent 3 hr per day 0.80 1.00 1.50
or hydraulic Over 3 hr up to and incl. 1.00 1.25 1.75
motor 10 hr per day
Over 10 hr per day 1.25 1.50 2.00

Multicylinder Occasional V2 hr per day 0.80 1.00 1.50

internal Intermittent 3 hr per day 1.00 1.25 1.75
combustion Over 3 hr up to and incl. 1.25 1.50 2.00
engine 10 hr per day
Over 10 hr per day 1.50 1.75 2.25

Single cylinder Occasional V2 hr per day 1.00 1.25 1.75

internal Intermittent 3 hr per day 1.25 1.50 2.00
combustion Over 3 hr up to and incl. 1.50 1.75 2.25
engine 10 hr per day
Over 10 hr per day 1.75 2.00 2.50
Source: Ref. 1.

expected since the enclosed drive standards reflect gear manufacturer’s field
experience and are based on practical as well as theoretical considerations.
A drawback of the enclosed drive standards is that they do not quantify
the effects of component quality or metallurgy in the rating procedures. A given
gearbox may contain gears with better metallurgical characteristics or closer
tolerances than another, yet both units could claim the same rating according to
existing standards. There is work proceeding to improve the standards. If an
application is considered critical, the user should determine the specific quality
and metallurgical characteristics required and identify them in the procurement
document rather than simply call for a unit complying with a particular
standard.
Very high speed gearing operating above 20,000 fpm pitch line velocity
requires analysis beyond that presented in industry standards. If a procedure
such as that outlined in the API standard is used to size a very high speed unit,
the gear diameters that result may become so large that centrifugal effects may
cause stresses and deflections that override the conventional strength and dura­
bility considerations. In some high-speed applications it is preferable to minimize
118 Gear Drive Systems

pitch line velocity at the expense of higher bending and compressive stresses.
These types of units must be carefully designed and incorporate the highest-
quality components with the best metallurgy available.

API STANDARD

The American Petroleum Institute standard [3], like the AGMA standards, is
widely used to procure gear units. There are two editions, the first issued in
1968 [8] and the second in 1977 [3]. The first edition rating method conforms
to the AGMA high-speed standard. The second edition rating method is more
conservative. Gear units are sized on the basis of a tooth pitting index called the
K factor. K-factor ratings are common in the industry and often used as a simple
rating index:

+ for external gear meshes, - for internal gear

meshes
where
K = tooth pitting index
Wt = transmitted tangential load, lb
F = net face, width, in.
PD = pinion pitch diameter, in.
Mg = gear ratio

Table 3.6 Material Index Numbers and Maximum L/d Valuesa

Maximum pinion
L/d
Gear hardness Pinion hardness Material Double- Single­
minimum minimum index number helical helical
223 Bhn 262 Bhn 130 2.4 1.7
262 Bhn 302 Bhn 160 2.3 1.6
302 Bhn 341 Bhn 200 2.2 1.5
352 Bhn 50 Rc (nitrided) 260 2.0 1.45
50 Rc (nitrided) 50 Rc (nitrided) 300 1.9 1.4
55 Rc (carburized) 55 Rc (carburized) 410 1.7 1.35
58 Rc (carburized) 58 Rc (carburized) 440 1.6 1.3

aL, net face width plus gap;

d, pinion pitch diameter.
Source: Ref. 3.
Gearbox Rating 119

The allowable K is the material index number shown in Table 3.6, divided by the
service factor:
v _ material index number
K. — —--------------------------
service factor
Also shown in Table 3.6 are maximum allowable face width/pinion pitch diam­
eter ratios. In the case of double helical gears the total face width L is the net
face width plus the gap between the helices. The bending stress is calculated as
follows:

WtPnd I-8 cos \p

St = (SF) ----------- psi

where

St = bending stress, psi

Pnd = normal diametral pitch, in.-1
F = net face width, in.
SF = service factor
* = helix angle, deg
J = geometry factor

The bending stress must not exceed the values shown in Figure 3.15.

50.000
CD THRU HARDENED TEETH AND CORE
5
3 40.000 |OF NITRIDED TEETH
Z
%$&+)%*&*)%$v%&&')%&&')%%&')%'&(*%%&')%%&r)
%
UJ 30.000
cc
fc
0
z 20.000
o
15,000

220 240 260 280 300 320 340 360 50 52 54 56 58

b r in e ll m a t e r ia l h a r d n e s s rockwell c

Figure 3.15 Bending stress allowables. (From Ref. 3.)

120 Gear Drive Systems

SAMPLE RATING CASE

Let us look at a specific application and see how the various rating procedures
compare. Consider the first mesh of a two-stage high-speed reducer connecting
a gas turbine with a generator having the following characteristics:
hp = 4500 hp
Pinion rpm = 14,500 rpm
Pinion pitch diameter = 4.625 in.
Number of pinion teeth = 37
Gear pitch diameter = 14.875 in.
Number of gear teeth = 119
Face width = 6 in.
Transverse pressure angle = 20.3439°
Helix angle = 11.0°
Transverse diametral pitch = 8.0
The gears are single helical, carburized, hardened to Rc 60 and ground to AGMA
Quality Class 12. The tangential load is
= 63,025(4500) x = 8 4 5 8 ]b
14,500 4.625

The pitch line velocity is

PLV = 14.500(4.625 ? _ = 17,557 fpm

12
Using Eq. (3.1), the bending stress is calculated:
Ka = 1.0; smooth operation is anticipated
Kv = 0.94; extrapolated from Figure 3.5

Ks = 1.0
Km = 1.32, assuming accurate gears and rigid mounting
J = 0.54 for the pinion
Figures 3.4b and 3.4c
J = 0.56 for the gear
„ 8 4 5 8 0 .0 ) x I x l f i ) = w
* 0.94 6 0.54
Gearbox Rating 121

Choosing a life factor Kj = 0.8, a temperature factor Kt = 1.0, and an allowable

material stress of 70,000 psi (Table 3.1), we have
K = 70,000(0.8) = 1 91
1 (1.0)(29,326)
which is a very low probability of failure.
If a reliability factor of 1.0 (a failure probability of 1 in 100) is chosen,
the allowable transmitted power on the basis of bending strength from Eq. (3.2)
is
P = 14,500(4.625X0.94) x _ 6_ x 0.54 x 70,000(0.8)
at 126,000(1.0) 1.32 1.0(8) 1.0(1.0)
= 8596 hp
To calculate the compressive stress, use Eq. (3.3):
Cc
I = _£
Mn
C = cos 20.3439° sin 20.3439° x 3522 = Q U 44
0 2 4.22
F 6
Mn = ——- = — = 0.558
Lmin 10-75
I = 0.2229
Ca = 1.0
Cv = 0.94
Cs = 1.0
Cm = 1.32
Cf = 1.0

S = 2300 / M M X - L ° _ X 1 ^ (L 0 1 = J
c V 0.94 4.625(6) 0.2229
Choosing a life factor Cj = 0.8, a hardness factor Ch = 1.0, a temperature
factor Ct = 1.0, and an allowable material stress of 225,000 psi (Table 3.3), we
have
C = 225,000(0.8X1.0) = x ?9
r 100,800(1.0)
122 Gear Drive Systems

Since Cr is lower than Kr, this gear mesh must be rated on the basis of durability
rather than tooth strength at the 4500-hp level.
If a reliability factor Cr of 1.0 (a failure probability of 1 in 100) is chosen,
the allowable transmitted power on the basis of durability from Eq. (3.4) is
= 14,500(6) 0.2229(0.94) x [225,000(4.625)
ac 126,000 1.0(1.32)(1.0)(1.0) L 2300
X 9_81 1 :Q) ] 2 = 14,360 psi
1.0(1.0)J
As the transmitted power increases, the bending stress increases linearly and the
compressive stress increases as the square root of the horsepower; therefore, at a
power point of approximately 6300 hp, the bending stress becomes critical. With
a reliability factor Cr of 1 .0 the bending stress is more critical than the compres­
sive stress and the allowable transmitted power is 8596 hp. The definition of Cr
as the reliability factor is relatively new. In the past it has been termed the factor
of safety.
On the basis of AGMA Standard 421.06 [2], the strength rating is
= 14,500(4.625)(0.695) __6_ 0.54 70,000(0.8)
at 126,000(1.0) 1.32 (1.0)(8) (1.0)(1.0)
= 6355 hp
The durability rating as per AGMA 421.06 is
c = 14,500(4.625)2(0.48) = j lg
1 126,000
C2 =4.2 (Figure 3.14)

C3 = I(Sac/Cp)2

I = 0.2229

(Sac/C p)2 = 6800 (Figure 3.12)

Pac = 7513 hp
The service factor at 4500 hp is
Gearbox Rating 123

On the basis of API Standard 613 [3 ], the strength rating is

c _ 8458(8.15) 1.8 cos 11°

---------- O r)
0.54

From Figure 3.15 the allowable bending stress number is 40,000 psi and the
service factor is
sp = 40,000(6)(0.54) = t Q6
8458(8.15)(1.8)(cos 11)
The API durability rating is

K = — - X — = 399
6(4.625) 3.22
The allowable material index number from Table 3.4 is 440, and in this case the
service factor is less than 1 .0:
399
SF = — = 0.91
440

SCORING

High-speed gearing (above 5000 fpm or 3600 rpm) operating with low-viscosity
lubricants is prone to a failure mode called scoring. In contrast to the classic
failure modes pitting and breakage, which generally take time to develop, scoring
occurs early in the operation of a gear set and can be the limiting factor in the
gear’s power capability. Scoring failures and the degree of scoring that may be
accepted are described in Chapter 12.

FLASH TEMPERATURE INDEX

The critical total temperature hypothesis (flash temperature index) [9] appears
to be the most reliable method of analysis presently used to predict scoring. It
states that scoring will occur when a critical total temperature, which is charac­
teristic of the particular combination of lubricant and gear material, is reached.
Tf = Tb + AT
where
Tf = flash temperature index, °F
Tb = gear blank temperature, °F
AT = maximum rise of instantaneous surface temperature in the tooth
mesh above the gear blank’s surface temperature
124 Gear Drive Systems

The gear blank temperature is difficult to estimate. It may be significantly

higher than the bulk oil temperature [10]. The heat transfer capability of the
gear must be considered in attempting to estimate this parameter. Often, the
blank temperature is approximated as the average of the oil temperature entering
and leaving the gearbox.
One form of the fundamental flash temperature index formula is [11]
CffW(Vi - V2)
Tf = Tb +
is /v ] + \N i) n /b J 2
where
Cf = material constant for conductivity, density, and specific heat
f = friction coefficient
Vx = rolling velocity of pinion at point of contact, fps
V2 = rolling velocity of gear at point of contact, fps
Bc = width of band of contact
W = specific loading, normal load divided by face width, lb/in.
For steel on steel gears, taking Cf as 0.0528, f as a constant 0.06, and adding a
term taking surface finish into account [ 10] , the following equation results
[12]:
/wte V4 50
T' = T>* \ t ) srnr z^ y
where
Wte = effective tangential load, lb
Fe = effective face width (use minimum contact length for helical
gears), in.
S = surface finish (after running in), rms
np = pinion rpm
Zt = scoring geometry factor

0.0175
Z. = --------- !=— -— J Note: Use absolute value of Zt
-

"1. 11
//44
*

(cos 0t)3/ r yg t
r * a

|_(ep +eg) J
where

ep = pinion radius of curvature, in.

eg = gear radius of curvature, in.
Np = number of pinion teeth (smaller member)
Gearbox Rating 125

Ng = number of gear teeth (larger member)

0t = pressure angle, transverse operating
The 50/(50 - s) term was developed by Kelly [10] in an experimental program
using gears with surface finish in the range 20 to 32 rms. For gears with surface
finish rougher than this range, if the computed value exceeds 3, a factor of 3
should be used. For gears with surface finishes finer than 20, the resulting com­
puted factor may be conservative.
The term Wte must be adjusted to allow for the sharing of load by more
than one pair of teeth. The following analysis, which modifies the tooth load
depending on the position of the gear mesh along the line of action, was devel­
oped by Dudley using spur gears of standard proportions [11]. If a more accu­
rate prediction of tooth load sharing is available to the reader, it would be
appropriate to use that analysis.

K = 1 + 1 9 " 9 lb
3 3 0L - 0Ld

K =1. 0

dL < 9 < dH

K I + 1 9° " 9
3 3 6o ~ 0H
9H <9 < 9o
2. Modified tooth profiles
a. Pinion driving
126 Gear Drive Systems

b. Pinion driven by gear:

1 + -6 B -
K
7 7 0L *
0LD < o < *L
K = 1.0
OS
IV

0L < en

K
= 6 K - e
7 00 - 0H
< e < 00
where
6 = any pinion roll angle, rad
6l d = roll angle at the pinion limit (form) diameter, rad
0L = roll angle at the lowest point of single tooth contact on the
pinion, rad
6H = roll angle at the highest point of single tooth contact on the
pinion, rad
6 0 = roll angle at the pinion outside diameter, rad
A modified tooth profile would be one that has tip and/or flank relief rather
than a true involute form.
The flash temperature index should be calculated at five specific points on
the line of action and at several additional points of contact. The five specific
points are:
1. Outside diameter of pinion
2. Highest point of single tooth contact
3. Pitch point (flash temperature rise will be zero since there is no sliding)
4. Lowest point of single tooth contact
5. Form (contact diameter) of pinion
Figure 3.16 is a typical plot of flash temperature rise along the line of action.
The most convenient way to generate a plot such as shown in Figure 3.16
is by the use of a computer program. By stepping through successive roll angles,
the flash temperature index can be calculated at many points. From Figure 3.16
it can be seen that there will be two peaks, one during the arc of approach
(pinion form diameter to pitch diameter) and one during the arc of recess
(pinion pitch diameter to outside diameter). To achieve the minimum flash
Gearbox Rating 127

SINGLE TOOTH CONTACT ZONE

Figure 3.16 Flash temperature rise along the line of action.

temperature index, the flash temperature rise in the arc of approach should be
equal to the rise in the arc of recess. An optimum tooth design can be achieved
by the use of long and short addendums. The computer program, starting with
standard addendums ( 1 /diametral pitch), automatically varies the pinion and
gear addendums in defined increments until the optimum flash temperature is
128 Gear Drive Systems

SCORING PROBABILITY (%)

Figure 3.17 Scoring probability versus flash temperature index. (From Ref. 12.)

obtained. With the resulting long and short addendum designs of this nature,
standard tooth thicknesses are no longer applicable. If standard tooth thick­
nesses were utilized, an unbalance of bending stresses between pinion and gear
would result. To optimize the bending stresses, the program enters a second
iteration procedure, which varies tooth thickness until bending stresses are
balanced.
Figure 3.17 presents the results of an aerospace industry survey correlating
scoring to flash temperature index. From the data used in the study, it was
assumed that a gearset with a calculated index of 276°F or less represents a low
scoring risk. A calculated index ranging from 277 to 338°F represents a medium
scoring risk where scoring may or may not occur, and a calculated index of
339°F and higher represents a high scoring risk. The data presented in Figure
3.17 reflects cases using Society of Automotive Engineers (SAE) 9310 steel
operating with military standard Mil-L-7808 or Mil-L-23699 synthetic oil. The
viscosity of these oils is approximately 4 to 6 cSt at 200°F and 18 to 30 cSt at
100°F. Mineral oils such as light turbine oils, used in high-speed industrial appli­
cations, are more viscous and therefore may possibly tolerate a higher flash tem­
perature index.
Gearbox Rating 129

The equation for the flash temperature index assumes a constant coeffi­
cient of friction of 0.06. If it is desired to calculate the coefficient of friction at
each point on the line of action, the following equation can be used [13]:
3.17 X 108
0.0127 log!
Mo VSV?/W

where
f = coefficient of friction
£Iq = absolute viscosity, cP
Vs = sliding velocity, ips
Vt = sum velocity, ips
W= specific loading, lb/in.
The equation breaks down at the pitch point, where the sliding velocity is 0.0
and the friction coefficient goes to infinity. Using a variable coefficient of
friction the flash temperature index formula becomes

Tf = + [Z t(np)V2]

0.2917

zt =
(cos 0t)3/ " epeg 1
_(ep + e g)J

SCORING C R ITE R IO N NUMBER

A simplified form of the flash temperature index is presented in Ref. 14. A scor-
ing criterion number is defined.
/wt\ 3/4 nn P1/2
Scoring criterion number =[— 1 pV4
\ e/ rd
where
Wt = tangential driving load, lb
Fe = contacting face width, in.
np = pinion rpm
Pd = diametral pitch
Table 3.7 gives scoring criterion numbers for various oils at various gear blank
temperatures. If the scoring criterion number is above the values shown in the
table, a possibility exists that scoring will be encountered. The gear blank tem­
perature can be taken as the average of the oil-in and oil-out temperatures.
130 Gear Drive Systems

Table 3.7 Critical Scoring Criterion Numbersa

Blank temperature (°F): 100 150 200 250 300
Kind of oil Critical scoring index numbers
AGMA 1 9,000 6,000 3.000
AGMA 3 11,000 8,000 5.000 2,000
AGMA 5 13,000 10,000 7,000 4,000
AGMA 7 15,000 12,000 9,000 6,000
AGMA 8A 17,000 14,000 11,000 8,000
Grade 1065 15,000 12,000 9,000 6,000
MH-Q-6082B
Grade 1010 12,000 9,000 6,000 2,000
Mil-0-6082B
Synthetic 17,000 14,000 11,000 8,000 5,000
(Turbo 35)
Synthetic 15,000 12,000 9,000 6,000 3,000
(Mil-L-7808D)
1/2
/ "t \
Scoring number =. (/ M
a “nP
p

Source : Ref. 14.

Ve — 1
/ Pd

M IN IM U M F IL M THICKNESS C R ITE R IO N

Scoring is a phenomenon that will occur when gears are operating in the boun­
dary lubrication regime [15] rather than with a hydrodynamic or elastohydro-
dynamic oil film separating the gear teeth. The film thickness can be calculated
[16-18] and compared to the combined surface roughness of the contacting
elements to determine if metal-to-metal contact is likely to occur. A criterion
used to determine the possibility of surface distress is the ratio of film thickness
to composite surface roughness:

x = ^nin
a
a = sJo2v + o\
where
X = film parameter
hmin = minimum oil film thickness, in.
Op = pinion average roughness, rms
og = gear average roughness, rms
Gearbox Rating 131

The “partial elastohydro dynamic” or “mixed” lubrication regime occurs if the

film parameter X is between approximately 1 and 4. At higher values, full hydro-
dynamic lubrication is established and asperity contact is negligible. Below a X of
1.0 there is a risk of surface distress.
The minimum elastohydrodynamic film thickness is calculated as follows
[16,17]:

2.65 G0'54 U0’70

H
\y 0*13

H = (film thickness parameter)

h 0 = minimum film thickness, in.

R = equivalent radius, in.
eDeg
R = —— - + external, - internal
ep ± e g
ep = pinion radius of curvature, in.
eg = gear radius of curvature, in.
G = aEr (materials parameter)
a = pressure viscosity coefficient, in.2/lb (Figure 3.18)

I = I f1-5' + [ : J A
2
E' 2 \ Ei E2 /
E' = 33 X 106 for steel on steel
5 1 = pinion Poisson’s ratio
82 = gear Poisson’s ratio
Ei = pinion Young’s modulus
E 2 = gear Young’s modulus
W'
W = ----- (load parameter)
ER
W' = Specific loading, lb/in.
Vi{\l + v2)mo
U
E'R
vx = wpep = rolling velocity of pinion at point of contact, ips
132
(x 105) (in.2/lb .)
PRESSURE VISCOSITYCOEFFICIENT Gear Drive Systems

TEM PERATURE (°F)

Figure 3.18 Pressure viscosity coefficient versus temperature.

V2 = wgeg = rolling velocity of gear at point of contact, ips

Wp = pinion angular velocity, rad/sec
wg = gear angular velocity, rad/sec
Mo = absolute viscosity, Reyns (lb-sec/in.2)
eZk
~ 6*9(10 )6
e = specific gravity (Figure 3.19)

Zk = kinematic viscosity, cSt (Figure 3.20)

Gearbox Rating 133

-5 0 0 50 100 150 200

TEMPERATURE f P )

Figure 3.19 Specific gravity versus temperature.

An extensive survey of aerospace power gears operating with synthetic lubri­

cants at high temperature revealed that calculated oil films were on the order of
0.000010 to 0.000020 in. With surface roughness on the order of 20 rms it can
be seen that these gears are operating with X less than 1.0, are in the boundary
lubrication regime, and are therefore prone to scoring problems.
Table 3.8 shows the results of a computer analysis of a high-speed gear set
with standard addendums. The flash temperature index is the maximum flash
temperature rise, 71°F, plus the gear bulk temperature, 160°F. The index of
231°F presents a low scoring risk (Figure 3.17), which could be slightly reduced
by optimizing tooth proportions. The calculated coefficient of friction is sig­
nificantly lower than the assumed valued of 0.06, with a corresponding lower
flash temperature rise. The calculated minimum film thickness is 0.000020
with a X term of 0.71, indicating operation in the boundary lubrication
regime.
134
(CENTISTOKES)
VISCOSITY
KINEMATIC Gear Drive Systems

TEMPERATURE (°F)

Figure 3.20 Kinem atic viscosity versus tem perature.

Gearbox Rating 135

Table 3.8 Scoring Analysis of a High-Speed Gearset

Design parameters
Pinion teeth: 25 Gear teeth: 85
Pinion speed: 12,223 rpm Horsepower: 281
Face width: 1.0 in. Helix angle: 0
Diametral pitch: 10 Pressure angle: 25°
Lubricant type: Mil-L-23699 oil Gear blank temperature: 160°F
Oil viscosity: 1.22 (10"6) lb-sec/in.2 Pressure viscosity coefficient:
Gear material: SAE 9310 steel 11.4 (10“5) in.2/lb
Surface finish: 20 /tin. rms

EHD film
Roll Friction Flash Flash thickness, Film
angle, coefficient, temperature temperature bmin parameter,
9 (deg) f rise, AT (°F)a rise, AT (°F)b (Min.) X
15.32 0.0 0 0 0 0.0
16.41 0.014 0 40 22 0.77
17.50 0.018 18 58 21 0.73
18.60 0.021 23 68 20 0.72
19.69 0.023 28 71 20 0.71
20.78 0.025 30 70 20 0.71
21.87 0.027 30 64 20 0.72
22.96 0.030 30 60 20 0.71
24.05 0.032 22 42 21 0.73
25.14 0.035 13 25 21 0.75
26.23 0.041 5 7 22 0.76
27.32 0.040 7 9 22 0.78
28.41 0.034 13 25 23 0.80
29.50 0.031 22 41 23 0.81
30.59 0.028 25 52 24 0.84
31.68 0.026 25 59 25 0.88
32.77 0.024 25 62 26 0.91
33.86 0.021 22 61 27 0.96
34.95 0.019 18 57 29 1.01
36.04 0.016 13 48 30 1.08
37.13 0.012 7 33 34 1.19
af is variable.
bf held constant at 0.06.
Source : Preventing Gear Tooth Scoring, Machine Design, March 20,1980, Penton/IPC.
136 Gear Drive Systems

SHAFT RATING

Three elements of shaft design are covered in this section:

Shaft stresses
Keyways
Splines

SHAFT STRESSES

Figure 3.21 gives maximum allowable stresses in a shaft due to torsion and bend­
ing. These allowable stresses provide for a stress concentration not exceeding 3.0

Figure 3.21 Allowable shaft stresses. (From Ref. 1.)

Gearbox Rating 137

and are based on a service factor of 1.0. It is assumed that over 1 million operat­
ing cycles will be accumulated. If the shaft is designed to a finite life of less than
1 million cycles, higher allowables can be used.
Let us review a simple example to demonstrate shaft design. The spur gear
shaft shown on Figure 3.22 transmits 500 hp at 10,000 rpm. The torque is there­
fore 3151 in .-lb and if the pitch diameter is 3.0 in., the tangential load is 2101 lb.
For a 20° pressure angle the separating load is 765 lb and the resultant radial
load is 2236 lb. The bearing reactions are 559 and 1677 lb, respectively.
To find the bending stress at section AA, use

where
M = bending moment, in .-lb
c = radius where stress is calculated, in.
I = moment of inertia of section, in.4 , for a hollow shaft
I = ( n /64) (O.D .4 - I.D 4)
The bending moment at section AA is
M = 1677(2.0) = 3354 in.-lb

I = — (1.54 - 1.24) = 0.147 i n 4

64
and
G _ 3354(0.75) _ iai10 .
S = ------ ^----- = 17,112 psi
b 0.147
138 Gear Drive Systems

To find the torsional stress, use

s . - ITr
t
I
where
T = torque, in.-lb
r = radius, in.
J = polar moment of inertia, in.4 ; for a hollow shaft
J = (tt/32)(O .D .4 -I.D .4)

At section AA:

J = — (1.5 4 - 1.24) = 0.293 in 4

32
and
q 3151(0.75) _ .
S = ------ ^ = 8065 psi
s 0.293
From Figure 3.21 it can be seen that the shaft hardness must be oyer 240 Bhn
for the stress levels calculated. The analysis given above for shaft bending and
torsion will yield satisfactory results for most applications. If it is desired to
minimize shaft weight or if a shaft has an unusual configuration, a detailed anal­
ysis combining tensile, shear, and compressive stresses must be conducted. Also,
stress concentrations must be calculated accurately. When refining a shaft analy­
sis in order to minimize shaft weight, deflections must also be considered, since
this factor may be more critical than stress.

KEYWAYS

Keyways must be analyzed for shear and compressive stresses as follows [6]:
2T
Ss
dwL
2T
Sc
dhi L
where

Ss = shear stress, psi

Sc = compressive stress, psi
T = shaft torque, lb-in.
Gearbox Rating 139

Table 3.9 Allowable Keyway Stresses

Allowable stress (psi)
Key material Hardness Bhn Shear Comp.
AISI 1018 None specified 7,500 15.000
AISI 1045 225-265 10,000 20.000
225-300 15.000 30.000
AISI 4140 310-360 20.000 40.000
Source: Ref. 1.

d = shaft diameter, in. (for taper shaftuse mean diameter)

w = width of key, in.
L = length of key, in.
hi = height of key in the shaft or hubthat bears againstthe key way; for
designs where unequal portions of the keyway are in the hub or
shaft, hi must be the minimum portion
Allowable keyway stresses are presented in Table 3.9.

SPLINE RATING

Standard SAE splines are rated on the basis of tooth shear and compressive
stress:

where
T = torque, in.-lb
D = spline pitch diameter, in.
Fe = effective face width, in.
Kjh = load distribution factor, Table 3.10
The compressive stress equation simply divides the tangential load per tooth by
the tooth bearing area:

Tangential load per tooth = ------------------------

D(number of teeth)
140 Gear Drive Systems

Table 3.10 Spline Load Distribution Factor (Km )

Misalignment (in./in.) Face width of splines

0.5 1.0 2.0 4.0
0.001 1.0 1.0 1.0 1.5
0.002 1.0 1.0 1.5 2.0
0.004 1.0 1.5 2.0 2.5
0.008 1.5 2.0 2.5 3.0
Source: Ref. 19.

Bearing area = Fe (tooth height)

Tooth height = 2 (spline addendum)
= 2_____________
denominator of diametral pitch
= 1____________
numerator of diametral pitch
pitch diameter
number of spline teeth
For instance, a 20-tooth, 20/40-diametral pitch spline has a tooth height of
0.5 in.
The shear stress equation divides the tangential load by the shear area at the
pitch line, which is

Shear area = --------------- ------------------------

2(numerator of diametral pitch)
Fe7r (pitch diameter)
2(number of teeth)
In the shear stress equation the assumption is made that only half the spline
teeth are in contact. Allowable shear stresses for splines are 50,000 psi for
hardened splines (Rc 60) and 40,000 psi for splines in the range Rc 33 to 38.
To determine the compressive stress allowable, the Sc calculated above is
modified by the following factors [19]:

ScK a
Sliding splines - Sc = ------
Lw
Gearbox Rating 141

Table 3.11 Spline Application Factor (Ka)

Type of load
Light Intermittent Heavy
Power source Uniform shock shock shock
Uniform 1.0 1.2 1.5 1.8
Light shock 1.2 1.5 1.8 2.1
Medium shock 2.0 2.2 2.4 2.8
Source: Ref. 19.

Table 3.12 Spline Wear Factor (Lw)

Number of Wear factor,
shaft revolutions Lw
10,000 4.0
100,000 2.8
1 million 2.0
100 million 1.0
1 billion 0.7
10 billion 0.5
Source: Ref. 19.

Table 3.13 Spline Fatigue Life Factor (Lf)

Life factor, Lf
Number of
torque cycles Unidirectional Fully reversed
103 1.8 1.8
104 1.0 1.0
10s 0.5 0.4
106 0.4 0.3
107 0.3 0.2
Source: Ref. 19.
142 Gear Drive Systems

Table 3.14 Allowable Compressive Stress for Splines

Allowable stress (psi)

Material Hardness Straight Crowned
Steel 160-200 Bhn 1,500 6,000
Steel 230-260 Bhn 2,000 8,000
Steel 33-38 Rc 3.000 12,000
Surface hardened 48-53 Rc 4.000 16,000
Case hardened 58-63 Rc 5.000 20,000
Source: Ref. 19.

Fixed or locked splines - §'c

where
Ka = application factor (Table 3.11)
Lw = wear factor (Table 3.12)
Lf = life factor (Table 3.13)
The life factor is based on the number of times the torque is applied and re­
moved in the expected life of the machine (number of starts). The wear factor is
based on the revolutions of the splines. Each revolution there is a back-and-forth
rubbing of the teeth, which causes wear. The value of S'c should not exceed the
allowable compressive stress numbers given in Table 3.14.

REFERENCES

1. AGMA Standard 420.04, Practice for Enclosed Speed Reducers or In­

creases Using Spur, Helical, Herringbone and Spiral Bevel Gears, American
Gear Manufacturers Association, Arlington, Va., December 1975.
2. AGMA Standard 421.06, Practice for High Speed Helical and Herringbone
Gear Units, American Gear Manufacturers Association, Arlington, Va.,
January 1969.
3. API Standard 613, 2nd ed., Special Purpose Gear Units for Refinery Ser­
vices, American Petroleum Institute, Washington, D.C., February 1977.
4. AGMA Standard 218.01, Rating the Pitting Resistance and Bending
Strength of Spur and Helical Involute Gear Teeth, American Gear Manu­
facturers Association, Arlington, Va., 1982.
5. AGMA Information Sheet 226.01, Geometry Factors for Determining the
Strength of Spur, Helical, Herringbone and Bevel Gear Teeth, American
Gear Manufacturers Association, Arlington, Va., August 1970.
Gearbox Rating 143

6. AGMA Gear Handbook 390.03, Gear Classification, Materials and Measur­

ing Methods for Unassembled Gears, American Gear Manufacturers Associ­
ation, Arlington, Va., January 1973.
7. Buckingham, E., Analytical Mechanics o f Gears, McGraw-Hill, New York,
1949.
8. API Standard 613, first ed., High-Speed Special-Purpose Gear Units for Re­
finery Services, American Petroleum Institute, Washington, D.C., August
1968.
9. Blok, H., Surface Temperatures Under Extreme Pressure Conditions, 2nd
work, Petroleum Congress, Paris, September 1937.
10. Kelley, B. W., A New Look at the Scoring Phenomena of Gears, SAE
National Tractor Meeting, September 1952.
11. Dudley, D. W., Practical Gear Design, McGraw-Hill, New York, 1954.
12. AGMA Information Sheet 217.01, Gear Scoring Design Guide for Aero­
space Spur and Helical Power Gears, American Gear Manufacturers Associ­
ation, Arlington, Va., October 1965.
13. Benedict and Kelley, Instantaneous Coefficients of Gear Tooth Friction,
ASLE Transactions 4, 1961, pp. 59-70.
14. Dudley, D. W., Gear Handbook, McGraw-Hill, New York, 1962, pp. 13-44,
13-45.
15. USAAMRDL-TR-75-33, Gear Tooth Scoring Investigation, Southwest
Research Institute, San Antonio, Tex., July 1975.
16. Dowson, D., Interdisciplinary Approach to the Lubrication of Concen­
trated Contacts, NASA SP-237, July 1969.
17. Dowson, D., The Role of Lubrication in Gear Design, British Institution of
Mechanical Engineers, 1970.
18. Zaretsky, E. V. and Anderson, W. J., How to Use What We Know About
EHD Lubrication, Machine Design, November 7, 1968, pp. 167-173.
19. Dudley, D., When Splines Need Stress Control, Product Engineering,
December 1957.
4
BEARINGS AND SEALS

BEARINGS

One of the first questions to be asked when reviewing a gear box design is: “How
are the shafts supported” ? Each shaft must be radially and axially located by
bearings and proper bearing design and application is at least as important to the
operation of the unit as the gears themselves. In fact, bearing problems are more
common in gear boxes than are gear failures. A basic choice when designing a
unit is which type of bearing to use. There are two general classes, one being
journal-type bearings (sometimes called fluid film or slider), the other being
rolling-element-type bearings (sometimes called antifriction).
Journal bearings are characterized by operating with a relatively thick oil
film between the rotating and stationary elements, the oil film being sheared by
relative sliding. Rolling element bearings include ball, roller, and needle bearings
with many design variations in each of these classes. The rolling elements are in
intimate contact with one another, the oil film being relatively thin compared to
journal bearings. In many applications it is not clear which configuration to use.
Let us look at various parameters as they relate to journal and rolling element
bearings.
1. Load capacity. As will be seen later in the chapter, rolling element
bearing fatigue life can be calculated on a statistical basis; therefore, in critical
applications where load is significant and/or speeds are high, finite lives are
predicted. Journal bearings are thought of as having infinite lives provided that
the loading is kept below a predetermined level. Therefore, in many applications
requiring long life, journal bearings are chosen since the calculated rolling
element life is considered insufficient. It is true that a journal bearing when

145
146 Gear Drive Systems

operating well can last indefinitely while a rolling element bearing eventually is
prone to fatigue failure; however, the L 10 bearing life concept is probably con­
servative and rolling element bearings quite often operate far longer than
predicted lives.
In many applications load is proportional to speed; however, if load is
applied at startup, rolling element bearings have an advantage over journal
bearings since a journal bearing must attain some speed before developing a
hydrodynamic oil film. In some cases journal bearings are externally pressur­
ized at startup so that the load is carried by this pressure rather than a hydro-
dynamic oil film. This is often done with heavy rotors such as large generator
or turbine shafts.
2. Speed. Journal and rolling element bearings are both speed limited
but for different reasons. As journal bearing speeds increase, the conjunction
zone becomes more turbulent and lubricant shearing increases the temperatures,
thereby decreasing the oil film thickness. Surface velocities at the bearing bore
of 15,000 ft/min are considered high, although operation to 30,000 ft/min has
been attained. Rolling element bearing speeds are limited by the centrifugal
forces generated between the balls or rollers and outer races, which can exceed
the capacity of the bearing. Rolling element bearing speeds are characterized by
the DN value (shaft diameter in millimeters times speed in rpm). DN values of
over 1 X 106 are considered high. Some gas turbine advanced applications
incorporate bearings operating at 3 X 106 DN; however, these bearings require
extensive development in such areas as lubrication and retainer design for proper
operation.
3. Lubrication. Rolling element bearings require significantly less oil
flow than journal bearings. As a result, they exhibit less power consumption and
heat generation. They also have lower starting torque requirements, particularly
at low temperatures where lubricant viscosity is high. Because of their large
operating oil films, journal bearings provide some damping in the mechanical
system, whereas rolling element bearings are stiff. Journal bearings are very
sensitive to dirt in the lubricant since the operating surface is much softer than
the hardened contact surfaces of rolling element bearings. Sometimes rolling
element bearings are chosen over journal bearings because pre- or postlubrication
can be eliminated. It is not necessarily true that journal bearings need to be
lubricated prior to startup, and many applications depend on a soft external
bearing layer such as babbitt to survive the initial cycles until an oil film forms.
For a given application, testing is the only positive way to determine if prelubri­
cation is necessary. Postlubrication is rarely necessary for either rolling element
or journal bearings, being used only if some high soak-back temperature will
exceed the stabilization temperatures of the bearing materials or cause oil left
in the bearing cavity to coke up.
Bearings and Seals 147

4. Cost. In very large quantities journal bearings can be produced at

lower cost than rolling element bearings. An example of this is crank shaft
bearings in the auto industry. In small quantities for special designs, journal
bearings are more expensive. The rolling element bearing industry has stan­
dardized on a series of bearing types and manufactures these in large quantities
as off-the-shelf items. Equivalent journal bearings for gearboxes, although some­
times listed in catalogs, are not mass produced and in the quantities purchased
are relatively expensive.
In many cases the cost of replacing a bearing and the downtime involved
is far more significant than the original cost. An example of such a case is a
gearbox in a critical process in a petroleum plant. In such an instance journal
bearings are preferred, for three reasons:
As discussed before, journal bearings, because the moving parts are separated by
a large oil film, can last indefinitely, whereas rolling element bearings are
subject to fatigue failures.
The types of failure encountered with journal bearings are less catastrophic and
easier to detect. For instance, a journal bearing that is wearing can be
monitored with proximity probes and replaced at a convenient time. A
spalling rolling element bearing is harder to detect, is introducing abrasive
debris into the lubrication system, and may fail abruptly as surface deteri­
oration progresses.
Journal bearings can be split, which enables the user to replace a defective
bearing without removing the shafts from the gearbox. This way the
couplings remain in place and a time-consuming realignment is not
required.
In a broad overview of when journal or rolling element bearings are used:
Journal bearings tend to be chosen for high-speed operation, where long life and
extreme reliability are desired and the bearing cost is small compared to the cost
of system downtime. Lower speed, less critical gearboxes tend to incorporate
rolling element bearings. In some sophisticated applications such as aircraft
turbine engines, rolling element bearings are used to take advantage of low oil
flow requirements, high efficiency, ability to start without oil and at low
ambient temperatures.
Once the choice between journal or rolling element bearings has been
made, a more detailed analysis of bearing selection and application is in order.
The next section will be a review of rolling element bearing fundamentals fol­
lowed by a consideration of journal bearings.

Rolling Element Bearings

Rolling element bearings are composed o f four elements: an inner race, an outer
race, a complement o f balls or rollers, and a ball or roller separator (sometimes
148 Gear Drive Systems

called a retainer or cage). Figure 4.1 illustrates bearing nomenclature. Usually,

the inner race rotates and the outer race is stationary, but there are applications
where the reverse is true or both races rotate. Ball bearings in addition to react­
ing radial loads can carry significant thrust loads, whereas roller bearings though
having greater radial load capacity than ball bearings can react only small thrust
loads. The greater roller bearing radial capacity is due to the fact that rollers are
in line contact with the races as compared to the point contact of ball bearings;
therefore, the radial load is spread over a greater area. The following paragraphs
describe the most widely used rolling element bearing configurations.
Deep Groove Ball Bearings This configuration, sometimes called the
Conrad type, is illustrated in Figure 4.2. This type of bearing is commonly used
in a wide variety of applications where either thrust or radial capacity or both
are required. The bearing is assembled by bunching the balls in one arc of the
circumference, which permits the inner and outer races to be located in their
proper positions. The balls are then separated around the circumference and
their retainer fitted in place. A number of retainer configurations are used, the
most common being a pressed steel design with the two halves riveted together.
Bearings and Seals 149

FILLING GROOVES

Figure 4.2 Deep groove ball bearing. A. Conrad type. B. Filling groove type.
(Courtesy of SKF Industries, Inc., King of Prussia, Pa.)

Because the load capacity is dependent on the number of balls, several design
variations are available which increase the ball complement. One such design is
the filling notch bearing, where a single notch is ground into both the inner and
outer race through which additional balls can be fed. For maximum strength and
high-speed operation a split inner ring bearing can be used where the inner race
is made of two sections. In addition to allowing the maximum complement of
balls to be assembled, a strong one-piece machined steel retainer can be used in
this configuration. It is also possible to split the outer race axially in order to
150 Gear Drive Systems

C O N TA CT ANG LE

ef

Figure 4.3 Angular contact bearing. (Courtesy of SKF Industries, Inc., King of
Prussia, Pa.)

assemble more balls. In the case of the split inner race design the balls do not
touch the split, so no reduction in fatigue life occurs. This is not the case with
the split outer race bearings.
Angular Contact Ball Bearings Angular contact bearings are designed for
applications with high thrust loads in one direction. Figure 4.3 shows the design
features, which include one high thrust shoulder on the inner and outer race on
opposite sides. There is a low shoulder opposite the high shoulder on each race;
thus the maximum number of balls can be snapped into place during assembly.
The contact angle between the balls and raceways is designed to be higher than
a deep groove bearing giving more thrust capacity. Angular contact bearings can
be used in pairs to provide rigid axial location and high thrust capacity in either
direction. These bearings can react a combination of thrust and radial loads but
are usually used when the thrust load is predominant.
Duplex Ball Bearings Duplex ball bearings are a pair that can be mounted
in four ways, as shown in Figure 4.4, to accommodate different loading condi­
tions and stiffness requirements. The inner and outer races are machined such
that there is a controlled axial relationship between the two. When the bearings
are mounted in the tandem configuration, the precise machining will enable the
bearings to share the thrust load equally. In this case the contact angles are
parallel. In the face-to-face configuration the contact angle lines converge
Bearings and Seals 151

--------21 S ........ -

D ------------- ------------ d

U 1w

FACE-TO-FACE

THREE BEARINGS IN TANDEM

Figure 4.4 Duplex bearing arrangements. (Courtesy o f SKF Industries, Inc.,

King of Prussia, Pa.)
152 Gear Drive Systems

Figure 4.5 Self-aligning ball bearing. (Courtesy of SKF Industries, Inc., King of
Prussia, Pa.)

inwardly. This design is used for heavy radial or combined radial and thrust loads
or for reversing thrust loads. The face-to-face arrangement allows the bearing to
accommodate a small amount of misalignment. The back-to-back configuration
has the same large load capacity as the face to face but adds angular rigidity to
the system and may be used where it is necessary to restrict misalignment or
shaft deflection. The contact angle lines diverge outwardly.
Self-Aligning Ball Bearings Figure 4.5 shows a self-aligning ball bearing
which can accommodate large amounts of misalignment. This is accomplished by
machining the outer race to a spherical contour; however, the large radius of
curvature reduces the bearing-load capacity. Another method of accommodating
misalignment is to make the outside diameter of the bearing a spherical surface
which then mounts into a spherical seat in the housing. The bearing is then free
to position itself.
Double Row Ball Bearings The double row ball bearing (Figure 4.6) con­
sists of one-piece inner and outer races each having two raceways and two ball
complements. This design provides heavy radial and thrust capacity in either
direction in an envelope somewhat smaller than two single-row bearings. The
contact angles can be either face to face or back to back depending on whether
it is desired to accommodate misalignment or provide a stiff mounting.
Ball Thrust Bearings Figure 4.7 shows a typical ball thrust bearing, which
is similar to a radial ball bearing except that the raceways are axial rather than
radial.
Bearings and Seals 153

Figure 4.6 Double row ball bearing. (Courtesy of SKF Industries, Inc., King of
Prussia, Pa.)

Figure 4.7 Ball thrust bearing. (Courtesy of SKF Industries, Inc., King of
Prussia, Pa.)
154 Gear Drive Systems

Single Row Single Row

Figure 4.8 Cylindrical roller bearings. (Courtesy of SKF Industries, Inc., King
of Prussia, Pa.)
Bearings and Seals 155

Cylindrical Roller Bearings These bearings, illustrated in Figure 4.8,

feature cylindrical rollers which run on cylindrical raceways. The roller length
is approximately equal to the roller diameter and the rollers are crowned to
relieve potentially high stresses at their ends. Roller retainers are positioned
either by the rollers or the inner or outer races and are usually two-piece con­
struction either riveted or screwed together. Figure 4.8 shows some of the
cylindrical roller bearing configurations available. On the left side are two float­
ing designs where either the inner or outer race has no shoulders and the shaft is
allowed to float axially in relation to the housing. Only radial loads can be trans­
mitted. This type of bearing is useful when the shaft must be allowed to move
axially due to thermal expansion or deflection due to loading. Cylindrical roller
bearings with one shoulder on one race and two shoulders on the other race
allow axial movement in one direction and can sustain some thrust load in the
other direction. Designs are also available with two shoulders on both the inner
and outer races and can sustain axial loads in both directions. In this case one of
the races is machined in two pieces for ease of assembly and inspection. It is
difficult to determine the amount of thrust load a cylindrical roller bearing is
capable of transmitting. The load is carried by sliding contact between the roller

Figure 4.9 Spherical roller bearing. (Courtesy of SKF Industries, Inc., King of
Prussia, Pa.)
156 Gear Drive Systems

ends and the race lips such that the action is more like a journal thrust bearing
than a rolling element bearing. The important parameters, therefore, are the
contact area, surface finish, surface geometry, sliding velocity, lubricant, and
operating temperature. In general, a roller bearing can be used to axially locate
a shaft when no large thrust loads are anticipated, but should not be expected
to react significant amounts of thrust.
Spherical Roller Bearings Figure 4.9 illustrates a typical spherical roller
bearing. There are two rows of rollers which run on a common raceway in the
outer ring, which has been ground to a spherical contour on the inner diameter.
The inner ring has two raceways ground at an angle to the axis of rotation; thus
the bearing is capable of reacting moderate thrust loads in either direction, in
addition to heavy radial loads. The spherical outer race enables the bearing to
accommodate some shaft misalignment. This type of bearing is suitable for
relatively low speed operation.
Tapered Roller Bearings The tapered roller bearing (Figure 4.10) is
designed such that lines extended from each tapered surface intersect at a
common point on the bearing axis. Because of the tapered races heavy loads in
both the radial and axial directions can be handled. As shown in Figure 4.10,
both single row and double row configurations are available. Tapered roller
bearings are usually limited to low-speed operation.
Needle Bearings As shown in Figure 4.11, needle bearings incorporate
rollers which are relatively long in relation to their diameter. Their major advan­
tage is high radial load capacity with a small radial envelope requirement.
Because of the large length-to-diameter ratio, needle bearings are very sensitive
to shaft misalignment.

Bearing Life Rating

Bearings are conventionally rated in terms of L 10 life, the life that 90% of a
group of bearings operating at a given set of conditions will complete or exceed.
A bearing failure is defined as the first occurrence of fatigue on one of the
rolling elements or on one of the raceways. Fatigue usually manifests itself in
the form of a spall. The L 10 life concept is based on extensive experimental data
and as with all fatigue data, there is considerable scatter; however, on the average
a plot of life versus percentage of bearings failed takes the form shown in Figure
4.12. In the figure it can be seen that the median life, the life that 50% of a
group of bearings will achieve, is five times L 10 life.
The term “rating life” or just “life” of a bearing has been standardized
on as the Lx0 life by organizations such as the Anti-Friction Bearing Manu­
facturers Association (AFBMA), the American National Standards Insti­
tute (ANSI), and the International Standards Organization (ISO). It is ?
Bearings and Seals 157

single row, tapered roller bearings

double row, tapered roller bearings

Figure 4.10 Tapered roller bearings. (Courtesy of SKF Industries, Inc., King of
Prussia, Pa.)
158 Gear Drive Systems

Figure 4 .1 1 Needle bearing. (Courtesy of SKF Industries, Inc., King of Prussia,

Pa.)

BEARINGS FAILED {%)

Figure 4.12 Relative bearing life versus percentage failed.
Bearings and Seals 159

reasonable compromise that provides reliable bearing service while meeting

practical economic requirements. In other words, bearings chosen using this
criterion will provide reasonably long service, yet not be overly expensive. It
must be remembered that the L10 life concept pertains only to fatigue of the
rolling elements and raceways. Failures due to other causes, such as wear,
excessive heat generation, retainer failure, and so on, are not covered by the
fatigue model and in many cases limit bearing life.
Extensive testing by the bearing manufacturers, together with analytical
studies, has established that the fatigue life of ball bearings is inversely propor­
tional to the third power of the load and the fatigue life of roller bearings is
inversely proportional to the 3.33 power of the load:

for ball bearings

for roller bearings

where
La = bearing life at condition A, cycles
LB = bearing life at condition B, cycles
Fa = applied load at condition A, lb
FB = applied load at condition B, lb
Let us assume the following:
Lg = 1 million cycles
Fg = load (lb) that will give a life of 1 million cycles = C
La = L10 = rating life, millionsof cycles
Fa = P = applied load, lb
From the above are generated the familiar equations

for ball bearings

3.33
X 106 for roller bearings

The basic load rating C for any bearing can be found tabulated in bearing manu­
facturers’ catalogs. If the L10 life is desired in hours:

hours for ball bearings

 BALLBEARINGS H U L L h n B tA H IIM u b

Speed Load Life ^) Life1) Speed Load Life1) Life1)

rpm Ratio Million Hours rpm Ratio Million Hours
C/P Revolu­ — 200
20 —3 tions I— 200 io n s

E~ 300
30-
— 400
■ 1 .0
40- -4 0 0
1.0 - 4 -1 .0
j — 500
50- -5 0 0 -2
~ 600
60- ~ 2 -6 0 0 — 700
3
70- r- 3 -7 0 0 ■4 — 800

80- •5 — 900
90-
§- 4
=_5
- 800
-9 0 0
100- ■6
H 0 0 0

100- ~Q
— F- 1000 18

pa :10 -
— 1 500
r1 0
■ 20 :
-1500
H 2 000
30 E
-2000 ■ 40 r
30 50
L_ 3 0 0 0
. 60 =
- 40
• 80 if
50
L 4 000
— 60 ■100
: 00 I_ 5 000
100 -4 0 0 0
- 200
= - 6 000
500 - -5 0 0 0
300 — 7 000
mo - L-200
6-f -6 0 0 0 - 400 — 8 000
I - 9 000
• 500
700 - :=- 300 - TOGO
7— 1 000- ■ 600 - 1 0 0 0 0
@00 - = - 400 -8 0 0 0 “
- 800
900- -9 0 0 0 1 200
8-i L.500
1000- ; “ 600
9 — . E l 800
■1Q000
: 1 0 0 0
— 15 0 0 0
1200
1 800 ■ 2 000 I
10 r- 1000 10- " ~ 20 0 0 0
2 000 “
1500- -15000 3 000 I
=-
1800 - 4 000 E
2000 5 000
- -20000 = - 30 0 0 0
- 6 000
3000
■ 8 000 E.
5- 4000 = - 40 0 0 0
■10 0 0 0
5000
3000- -30000
- 0060 5 000 - 50 0 0 0

E^eooo 6 000 _ - 20 000

4000-
-10000 - 40000
7 000 - 30 000
r - 60 0 0 0

j- 70 0 0 0
5000 - -5 0 0 0 0 8 000 _ - 40 0 0 0 — 80 0 0 0
9 000- 1 - 90 0 0 0
-20000 ■ 50 0 0 0
6000- - 60000 10 000 - ■ 60 000 - 1 0 0 000
7000 - 30 00 0 -70000 - 80 000

8000 - -8 00 0 0 : 1 0 0 0 0 0
= - 40000
9000- 5 0 00 0 -9 00 0 0 15 000- r 150 000
10000 - — 60000
“ 100000
200 000
20 000 - r 200 000

- 150000 E
15000 ~
30 000 -J *-300 000
20000 r- 200000

30000 -3 L 300000

')This Life is expected to be exceeded by 90% of the

bearings.

Figure 4.13 Bearing life nomogram. (From Ref. 1.)

Bearings and Seals 161

3. 33
16,667
hours for roller bearings

where N is the inner or outer race net speed in rpm. The nomograms in Figure
4.13 represent the equations above and are often used to calculate bearing lives.
If the operating load P, operating rpm N, and bearing capacity C are known, a
straight line can be drawn that intersects the left scale at the operating speed
and the C/P scale at the calculated value of C/P. Where this line intersects the
hours scale is the rating life in hours. If the desired rating life, rpm, and applied
load are known, a straight line drawn between the speed and life axes will
intersect the C/P axis and the required bearing capacity can be calculated.
The L 10 life concept implies a reliability of 90% with regard to surviving
the rating life. In some critical applications reliability numbers greater than
90% are desired. For instance, to achieve a 97% reliability, an L 3 life, the life
that 97% of a group of bearings will reach or exceed at a given set of operating
conditions, must be calculated. Figure 4.14 gives an estimate of a life adjustment
factor corresponding to increased reliability. The factor is based on experimental
data and is used as follows:
Ln = aq L 10
where
oq = life adjustment factor
Ln = fatigue life expectancy for other than 90% reliability
For example, L 3 corresponds to 97% reliability and the factor , from Figure
4.14, is 0.425. If the calculated L10 life of a bearing is 10,000 hr, the L 3 life is
L3 = 0.425(10,000) = 4250 hr
Figure 4.14 indicates that there is no probability of fatigue failure at lives below
5% of the L10 rating life. In other words, in the example above the bearing is
100% reliable up to 500 hr of operation.

Rating Life Adjustments

Over the past years substantial increases in the fatigue endurance of rolling
element bearings have been made possible by improvements in bearing design,
materials, processing, and manufacturing techniques. Also, investigation into
effects of high-speed operation and misalignment has generated new insights
into their effect on load rating. Using the AFBMA L10 life method for deter­
mining fatigue life and the bearing load ratings in the various manufacturers’
catalogs to calculate the basic L10 life, an expected bearing life LA can be cal­
culated as follows:
 0.03 0.06 0.1 0.2 0.3 0.4 0.6 0.9 1 2 3 4 6 8 10
Probability of Failure, F (%)

Figure 4.14 Life adjustment factor for reliability. (From Ref. 1 .)

La = (D)(E)(F)(G)(H)L 1o
where D through H are life adjustment factors. This equation, developed by the
Rolling Elements Committee of the Lubrication Division of the American
Society of Mechanical Engineers and published as Ref. 2, explains the life adjust­
ment factors in detail. The following paragraphs summarize the work.
Material Factor D The predominant material for rolling element bearings
is AISI 52100. AISI is the American Iron and Steel Institute designation for
steels of various specific chemistry. The basic bearing dynamic capacities which
are presented in catalogs are based on air-melted 52100 steel through-hardened
to Rc 58 minimum. The mathematical model to define bearing load capacity
evolved in 1949 and since then the bearing steels have been improved such that
they are more homogeneous with fewer impurities. A materials factor D of 2 is
suggested for currently available steels. Case-hardened materials, used in tapered
roller bearings and other applications, have also improved over the years, but
insufficient data are available to recommend a materials factor.
Bearings and Seals 163

Processing Factor E The processing considered here is concerned mainly

with the melting practice. Air-melt material is considered the baseline and is
assigned an E factor of 1. Vacuum-melted material has an E factor of 3.
Although standard catalog bearings are not necessarily vacuum melt, quite often
this processing is used and the factor can be taken advantage of.
Lubrication Factor F Bearing life has been found to be significantly
affected by the thickness of the lubricant film developed between the contacting
elements. Film thickness is affected mainly by speed and lubricant properties at
the operating temperature with higher fatigue lives obtained at high speeds or
with higher-viscosity lubricants. Conversely, if a bearing operates with poor
lubrication film formation due to low speed or insufficient lubricant viscosity,
the life predicted from the catalog rating may not be achieved. The oil film
between rolling elements of a bearing has been found to be in the elastohydro-
dynamic regime and can be calculated according to the methods outlined in
Ref. 3 for roller bearings and Ref. 4 for ball bearings. The calculated oil film is
compared to the composite roughness of the contacting elements to form what
is referred to as the lambda ratio:
h
X
o
where
X - lambda ratio
h = oil film thickness, in.
o = composite surface roughness of the rolling element surfaces, in.
o = y/a\ + o\
where
<7l- rms surface finish of body 1 , in.
02 = rms surface finish of body 2, in.
The lubrication factor F is proportional to the lambda ratio and varies from
approximately 0.2 to 2.8. As an example, let us assume that the balls and race­
way of a bearing have rms surface textures of 0.000010 in. The composite
surface roughness will be 0.000014 in. If the calculated lubricant film thickness
is also 0.000014 in., the lambda ratio is 1. The lubrication factor, F in this case
is approximately 1 .
Speed Effect Factor G At high speeds the load at the outer race is in­
creased because the balls or rollers are subject to centrifugal forces and therefore
the fatigue life is reduced. Also, the load distribution on the rolling elements and
the zone of loading can change because of centrifugal growth and thermal
164 Gear Drive Systems

0.05 0.1 0.2 0.5 1 2 5 10

VJV^

Figure 4.15a Fatigue life adjustment factor a23 for bearing materials and lubri­
cation. Lubricant viscosity in application = v\ required lubricant viscosity = vx.
(From Ref. 5.)

expansion. High-speed bearings must therefore be analyzed using sophisticated

computer models which can take all the varying parameters into account, and
bearing manufacturers have such programs available for users. In general, cen­
trifugal effects will become significant at DN (bore in millimeters times speed
in rpm) values above approximately 0.5 X 106 and certainly bearings operating
above 1.0 X 106 DN should be carefully analyzed.
Misalignment Factor H The effect of misalignment on roller bearings is
to concentrate the load on one end and therefore reduce the fatigue life. Mis­
alignment can occur due to deflection under load or manufacturing errors in
the bearing housing or shaft. The following table provides some misalignment
limits beyond which reduction in bearing life may be expected.
Bearings and Seals

o
C

o
3oQ.
C/3
cin

Pitch diameter (dm) mm

Figure 4.15b Minimum required lubricant viscosity. (Bearing bore + OD) -r 2

= dm; required lubricant viscosity for adequate lubrication at the operating
temperature = v x. (From Ref. 5.)

Allowable Misalignment
Radians Minutes

Cylindrical and tapered roller bearings 0.001 3-4

Spherical bearings 0.0087 30
Deep groove ball bearings 0.0035-0.0047 12-16
aBased on general experience as expressed in manufacturers’ catalogs.
Source: Ref. 2.

When using roller bearings on flexible shafts the misalignment due to load should
always be calculated, since this factor may be life limiting.
166 Gear Drive Systems

We can summarize the section on rating life adjustments by noting that

today’s bearings, assuming good lubrication, can achieve fatigue lives signifi­
cantly greater than those calculated from catalog ratings. Three- to sixfold life
improvement ratings are possible.
One bearing manufacturer’s method of presenting these improvements is
outlined in [5]. A combined life adjustment factor designated a 23 replaces the
material and lubrication factors described in the preceding section. Figure 4.15a
is a plot of the a 23 factor that can be expected for a particular viscosity ratio
(p/vi) and bearing material combination. The viscosity ratio is the ratio of the
actual lubricant viscosity (V), at the operating temperature to the minimum
value required (iq), which can be found from Figure 4.15b. Figure 4.15b
presents the minimum viscosity at the operating temperature required for a given
bearing pitch diameter/speed combination. Some common lubricant viscosities
at various temperatures can be found in Figures 3.20 and 5.2.
Figure 4.15a is divided into two bands, one for vacuum-melted and one for
standard bearing steels. Under comparable operating conditions, certain bearing
types (e.g., spherical roller bearings, tapered roller bearings, and spherical roller
thrust bearings) normally have a higher operating temperature than other types,
such as deep groove ball bearings and cylindrical roller bearings. Therefore, the
upper boundary of either band is generally used for ball and cylindrical roller
bearings and the lower boundary is used for tapered and spherical roller bearings
[5].

Simultaneous Radial and Thrust Loads

Bearings frequently must carry a combination of radial and axial loads. To calcu­
late an L 10 life, the loading combination must be converted into an equivalent
load which would be the constant stationary radial load which if applied to a
bearing with rotating inner ring and stationary outer ring would give the same life
as that which the bearing would attain under the actual conditions of load and
rotation. The general equation for the conversion applying to rotating bearings is
P = XFr + YFa
where
P = equivalent load, lb
Fr = actual constant radial load, lb
Fa = actual constant thrust load, lb
X = radial factor
Y = thrust factor
The information required to calculate the X and Y factors for a specific bearing
is tabulated in bearing manufacturers’ catalogs.
Bearings and Seals 167

Bearing Selection Example

Figure 4.16 shows a typical page from a bearing catalog. The AFBMA has stan­
dardized on bore size, outside diameter, and width of ball and roller bearings and
these dimensions are commonly given in millimeters. For instance, in Figure
4.16 a 6020 bearing has a 100-mm bore, a 150-mm outside diameter, and a
24-mm width. This size bearing will be found in any manufacturer’s catalog and
although the internal geometry will vary, externally all 100 X 150 X 24 mm
bearings will be geometrically interchangeable. Let us assume that the 6020
bearing carries a radial load of 1000 lb and a thrust load of 500 lb at a constant
inner ring speed of 2500 rpm. Lubrication is oil jet with a 200 SUS viscosity
mineral oil. The bearing material is 52100 air melt. The radial and thrust load
are combined into an equivalent load using the procedure outlined in Figure
4.16:
Fr = 10001b
Fa = 5001b
V = 1 . 0 (inner ring rotating)
CQ = 9410 lb(basic static load rating)

— = 0.053
Co
e = 0.26
Y = 1.71
F„
= 0.5 (larger than e)
V(Fr)
X = 0.56
P = 0.56(1)(1000) + 1.71(500) = 14151b
C = 10,400 lb (basic dynamic load rating)
C
7.35
P
= 16^67 /C \ 3 = 2647 hr
2500 \Pf
The bearing pitch diameter is

PD = ° g - t bore = “ P -IM = 125mm

 Basic
Preferred Basic D yn a ­
Shoulder Balls Static mic Approx.
Diameter Brg. Load Basic
Nominal Bearing Dimensions M ax. Wt. Load Speed
in. Rating Brg.
Fillet lb Rating Limit2) No.
d D B F M S H T Radius1) Co C rpm

M BER mm mm in . mm in. in. in.

Shaft Housing No. Diam. lb lb
in. in. in. in. in. in.
6©00 10 .3937 26 1.0236 8 .3150 .012 .472 .945 .044 7 3/l6 440 790 28000 6000
6001 ' 12 .4724 28 1.1024 8 .3150 .012 .551 1.024 .049 8 3/6 500 880 25000 6001
6002 : 15 .5906 32 1.2598 9 .3543 1,187 1 7A6 .120 .078 .042 .012 .669 1.181 .068 9 3A 565 965 21000 6002
6003 17 .6693 35 1.3780 10 .3937 .012 .748 1.299 .089 10 625 1040 19000 6003

6004 20 .7874 42 1.6535 12 .4724 .024 .890 1.457 .16 9 Va 1000 1620 16000 6004
6005 25 .9843 47 1.8504 12 .4724 .024 1.087 1.654 .18 10 Va 1110 1740 14000 6005
6006 30 1.1811 55 2.1654 13 .5118 .039 1.346 1.929 .26 11 %2 1550 2290 12000 6006
6007 35 1.3780 62 2.4409 14 .5512 2.347 2 % .143 .078 .065 .039 1.543 2.205 .35 11 5A6 1910 2760 10000 6007

6003 40 1.5748 68 2.6772 15 .5906 2.552 2”jU .159 .094 .065 .039 1.748 2.441 .43 12 2090 2900 9200 6008
6009 45 1.7717 75 2.9528 16 .6299 .039 2.008 2.717 .54 13 2730 3630 8400 6009
6010 , 50 1.9685 80 3.1496 16 .6299 3.024 3% .159 .094 .065 .039 2.205 2.913 .58 14 "/& 2940 3770 7700 6010
6011 | 55 2.1654 90 3.5433 18 .7087 3.417 3514i .204 .109 .095 .039 2.441 3.268 .85 13 % 3820 4890 6900 6011

6012 60 2.3622 95 3.7402 18 .7087 .039 2.638 3.465 .92 14 % 4110 5090 6500 6012
6013 . 65 2.5591 100 3.9370 18 .7087 3.811 43/U .204 .109 .095 .039 2.835 3.661 .96 15 % 4410 5280 6100 6013
6014 70 2.7559 110 4.3307 20 .7874 .039 3.031 4.055 1.33 14 % 5480 6580 5500 6014 •
6015 : 75 2.9528 115 4.5276 20 .7874 4.402 4% .204 .109 .095 .039 3.228 4.252 1.41 15 % 5870 6830 5200 6015
6016 80 3.1496 125 4.9213 22 .8661 .039 3.425 4.646 1.87 14 '%2 7030 8240 4800 6016
6017 85 3.3465 130 5.1181 22 .8661 4.930 5V% .218 .109 .109 .039 3.622 4.843 1.96 15 l %2 7540 8560 4600 6017
6018 90 3.5433 140 5.5118 24 .9449 .059 3.898 5.157 2.56 14 8790 10000 4300 6018
6019 95 3.7402 145 5.7087 24 .9449 .059 4.094 5.354 2.67 15 9410 10500 4100 6019
6020 i 100 3.9370 150 5.9055 24 .9449 5.718 6% .250 .141 .109 .059 4.291 5.551 2.76 15 % 9410 10400 3900 6020
6021 105 4.1339 160 6.2992 26 1.0236 .079 4.528 5.906 3.51 15 % 11500 12500 3700 6021 •
6022 110 4.3307 170 6.6929 28 1.1024 6.443 73/, .261 .141 .120
6 .079 4.724 6.299 4.32 14 % 12900 14200 3500 6022
6024 120 4.7244 180 7.0866 28 1.1024 6.837 7% .261 .141 .120 .079 5.118 6.693 5.34 15 % 13800 14700 3200 6024

6026 130 5.1181 200 7.8740 33 1.2992 .079 5.512 7.480 8.16 15 % 17600 18400 2900 6026 •
6028 140 5.5118 210 8.2677 33 1.2992 .079 5.906 7.874 8.62 16 % 18800 19000 2700 6028
6030 150 5.9055 225 8.8583 35 1.3780 .079 6.378 8.386 10.6 16 Vb 21800 21800 2500 6030

C _ basic dynamic load rating from table J j J is smaller than or equal to e use X = 1 and Y = 0
When
P equivalent load (from formula below) VFr\is greater than e use X = 0 .5 6 and Y from table
Fa _
P = XVFr + YFa 0.014 0.028 0.056 0.084 0.11 0.17 0.28 0.42 0.56
Co"
where X = a radial factor given below e = 0.19 0.22 0.26 0.28 0.30 0.34 0.38 0.42 0.44
V = a rotation factor = 1.0 for inner ring) rotating in rela-
Y = 2.30 1.99 1.71 1.55 1.45 1.31 1.15 1.04 1.00
= 1.2 for outer ring) tion to the load
Y = a thrust factor given below C0 = basic static load rating from table.
Fr = the rad ial load, calculated
Fa = the thrust load, calculated 2)This refers to oil lubrication and moderate load. With grease lubrication
it is generally not practical to use speeds higher than % of those shown.
e is a reference value given in the table.

Figure 4.16 Typical bearing catalog data; series 60 single row, deep groove ball bearings. (Courtesy of SKF Industries, Inc.,
King of Prussia, Pa.)
Bearings and Seals 169

From Figure 4.15b the minimum required viscosity vx is 8 cSt. The 200
SUS oil viscosity v is converted to 40 cSt using the scales on Figure 4.15b. Note
that the relationship of SUS to centistokes is not linear; therefore, the v\vx ratio
used in Figure 4.15a must be calculated in centistokes. The v\vx ratio is 40/8 = 5
and from Figure 4.15a the life adjustment factor a23 is 4 for a standard steel.
The L 10 life, therefore, is four times the calculated catalog life of 2647 hr. If a
vacuum-melted steel is used, the life adjustment factor can be as high as 7. The
bearing DN value is 0.25 X 106 ; therefore, speed effects will not significantly
downgrade the life rating.

Permissible Speeds

Note that the catalog listing includes a speed limit for the 6020 bearing, 3900
rpm. This limit is based on permissible lubricant operating temperature and is set
somewhat below 250°F. Speed limits listed in bearing catalogs are not absolute
limits and by using proper design practices significantly higher operating speeds
are possible. The areas critical in high-speed operation are lubrication, cooling,
retainer design, and centrifugal effects.

Temperature Effects

When operating bearings at high temperatures, two potential problem areas are
phenolic retainers and steel stability. Phenolic retainers are temperature limited
and should not be used when operating temperatures approach 250°F. The
standard bearing steels are stabilized at 250°F and will change dimensionally if
operated at higher temperatures. It is possible to procure bearings stabilized at
temperatures above 250°F; however, this will reduce hardness and fatigue life.
It is also possible to procure high-hot-hardness steels such as M50, which have
excellent fatigue properties at temperatures up to 500°F.

Static Load Rating

The life load relationship would indicate that at zero speed a bearing has infinite
load capacity. Of course, this is not so. One could consider the limiting static
load, the load at which the bearing will fracture; however, at much lower levels
of loading, permanent deformations develop in the surfaces of the contacting
elements. The static load rating C0 is dependent on the magnitude of permanent
deformation that can be allowed. Experience has shown that when deformations
become as large as 1 / 10,000 of the diameter of the rolling element, objection­
able noise and vibration occur during subsequent rotation. The static load
rating given in Figure 4.16 is the load that corresponds to this magnitude of
permanent deformation. The fracture load is approximately eight times the
C0 value.
170 Gear Drive Systems

Stationary bearings subject to vibration may exhibit pitting at very light

loads. This is a form of fretting corrosion and can affect the fatigue life of
bearings that alternately rotate and are stationary, such as those used on over­
running clutches.

Prorating Bearing Loads

To calculate bearing L i 0 life an equivalent load is required; however, bearings

often operate under a varying load schedule. For instance, a typical operating
schedule for a bearing application follows:

Load point Time (min) Load (lb) rpm

1 7 300 1000
2 2 500 2000
3 1 700 3000

To find an equivalent load for the C/P term, the schedule can be prorated as
follows:

where
PE = equivalent load, lb
Nj = number of cycles at each load point
Pi = load at each point, lb
Nt = total number of cycles
For the example:
= C 7(1000)(300)3 + 2(2000)(500)3 + 1(3000)(700)3 1 1/3
E L 7
(1000) + 2(2000) + 1(3000) J
= 4971b
Each 10-min cycle consists of 14,000 revolutions. If the C/P value is calculated
to be 2, the life according to Figure 4.13 will be 8 million revolutions or 571
ten-minute cycles. The equivalent load is sometimes called the cubic mean load
and the exponent used is the same as for the life formula: 3 for ball bearings and
3.3 for roller bearings. When using this method of prorating, the difference
between the equivalent loads obtained with the two exponents is usually
negligible; therefore, the exponent 3 is used for both ball and roller bearing
calculations.
Bearings and Seals 171

Bearing Dimensions and Tolerances

The Antifriction Bearing Manufacturers Association has arrived at standardized

dimensions for rolling element bearings. These dimensions are the bore size, the
outside diameter, and the width, conventionally given in millimeters. The
variety of different sizes is limited so that bearings can be produced in large
quantities, thereby satisfying the user’s need for economical yet high-quality
bearings. The basic dimension in the system is the bearing bore, which defines
the shaft size upon which it is assembled. For a given bore a variety of outside
diameters and widths are available in increments such that requirements for load
capacity or envelope dimensions are satisfied. Through the dimensioning system,
bearings from different manufacturers or of different types can be interchanged.
It must be remembered, however, that although the bore, outside diameter, and
width of two bearings are the same, the internal geometry may be quite
different. This can affect the bearing capacity and operation. Manufacturers also
offer a variety of options, such as integral seals for grease-packed bearings, snap
ring grooves, and so on.
The Antifriction Bearing Manufacturers Association has set up a series of
quality classes known as grades ABEC-1, 3, 5, and 7. ABEC-1 is the standard
quality to which catalog bearings conform. The higher numbers correspond to
better quality: that is, smaller tolerances on the bore, outside diameter, and
width dimensions, as well as closer control of eccentricities, parallelism, and
squareness. Also, the higher classes will have better raceway surface texture and
the variation in ball or roller size within a bearing will be progressively less.
Grade 3 bearings are selected from the standard grade 1 production by
inspection. Grades 5 and 7 are manufactured separately with higher-precision
processing. The higher-accuracy grades are used when there are requirements for
especially smooth operation with low noise and vibration. An example of such
an application is a machine tool spindle. Also, high-speed bearing applications
should incorporate more precise bearings. Gearbox shafts operating above 3600
rpm or 5000 fpm peripheral speed should be mounted on ABEC-5 bearings.

Internal Clearance

The internal clearance of a bearing is defined as the total distance one bearing
ring can move in relation to the other ring in a radial direction under no load.
Ideally, bearings should operate with little or no radial clearance; however, this
condition is difficult to attain. The designed bearing clearance must take into
account the reduction in clearance if one or both of the raceways is mounted
with a press fit. Pressing on an inner bearing ring will reduce the internal clear­
ance 50 to 80% of the amount of the interference fit. Also, in operation the
inner ring will tend to be at a higher temperature than the outer ring. The outer
ring runs cooler since heat is more easily dissipated through the bearing housing
172 Gear Drive Systems

to the atmosphere than from the inner ring to the shaft. This temperature dif­
ferential means the inner ring expansion will reduce the bearing clearance.
Catalog bearings have standard clearances which are set such that the bearing
will operate properly when one of the raceways is mounted with a standard press
fit. For unusual conditions bearings are available with internal clearances
smaller and greater than standard. Bearing clearance will increase with bearing
size and the clearances, together with recommended press fits, are tabulated in
bearing catalogs.

Shaft and Housing Tolerances

Possible fits of inner races on shafts and outer races in housings range from loose
to transition to tight. One of the most important factors in determining suitable
fits at the bore and outside diameter is the relationship of the load to the bearing
rings. The load can be either stationary or rotating with respect to the rings. For
instance, in a gear shaft system using bearings with rotating inner rings the load
will always act in the same direction; therefore, the inner ring rotates in relation
to the load and all points on the inner race come under load each revolution. If
the inner ring were loose on the shaft, the relative motion of the load would
cause the ring to creep around the shaft, creating wear, fretting corrosion, and
the possibility of crack initiation. In the case of a stationary load and rotating
inner ring, therefore, the ring must have a tight fit on the shaft such that no
clearance exists and none can be developed by the action of the load. The higher
the load, the greater the press fit required. The outer ring, which in this case is
stationary with respect to the load, can be slightly loose in the housing to ease
assembly and disassembly. Clearance should be minimal, for the following
reasons:
1. A large clearance would allow the ring to cock at assembly.
2. A large clearance would reduce the centering ability of the bearing.
3. The outer ring can deform under load if not securely assembled in the
housing.
If the outer ring rotates and the load is stationary, the outer ring should be
assembled with a press fit to keep it from spinning. Quite often in shaft systems
there is a rotating load superimposed on stationary loads due to unbalance. This
rotating load tends to make loose-fitting outer races creep. A solution to this
problem is to restrain the outer ring from rotating with a pin or other locking
device.
The bearing shaft ideally should have a minimum hardness of Rc40 and be
ground to size. If these requirements are not practical in a particular application,
a tighter fit than normal should be used. In many applications bearing inner rings
are clamped axially. The axial restraint should not be counted on to restrain the
bearing from creeping and should not be used in place of a press fit.
Bearings and Seals 173

Bearing Mounting

The bearing arrangement must offer radial support and axial location. Also, it
must accommodate thermal expansion and deflections due to load. Figure 4.17
shows three common design solutions for the support of a gear shaft. Figure
4.17A illustrates a spur gear that generates no axial loads supported by two
roller bearings. The shaft is axially located by the bearing outer race shoulders
and one shoulder on each inner race. It is extremely important in this type of
design to allow sufficient endplay to accommodate thermal expansion of the
174 Gear Drive Systems

shaft. A detailed tolerance stack-up must be made to ensure endplay under the
worst dimensional conditions. A potential problem with this type of design at
high speeds is bearing skidding. This condition occurs when the bearings are
unloaded and do not have sufficient tractive force to keep them rolling. Skidding
will result in scoring and wear and may progress to complete bearing failure.
Some solutions for skidding problems are:
1. Reduction in bearing clearance such that the bearing always operates with a
slight preload giving sufficient tractive force.
2. Use of out-of-round outer races which impart a constant load on the bearing
in the zone where the rollers are pinched.
3. Reduction of forces that tend to retard rolling. In some cases lubrication is
excessive. Retainer design can alleviate the problem.
Figure 4.17B has a ball bearing which can react thrust in either direction on one
end of the shaft and a roller bearing on the other. The roller bearing has no
shoulders on the inner race and is free to accommodate thermal distortions. In
this case the gear might be helical. Care should be taken that the combination of
thrust and radial load on the ball bearing does not result in insufficient life.
Figure 4.17C illustrates a two-ball-bearing arrangement where the bearings are
preloaded by a spring or wave washer. This arrangement has the advantage that
at full speed and light load the spring ensures some bearing preload and the
chance of skidding is minimized. For some axial distance the spring rate is suffi­
ciently linear such that thermal expansion is accommodated and does not affect
bearing loading.
For heavy combined radial and thrust loads the arrangement shown in
Figure 4.18 can be used. The relief around the outside the diameter of the ball

Figure 4.18 Mounting configuration for combined radial and thrust loads.
Bearings and Seals 175

bearing ensures that all radial loads are reacted by the roller bearings and the ball
bearing reacts only thrust. The rollers are free to move axially to compensate for
temperature effects.

Tapered Roller Bearing Mounting

Figures 4.19 and 4.20 show two common ways to mount bevel gears in tapered
roller bearings. Figure 4.19 illustrates a straddle mount where the gear is sup­
ported on either side. In the case shown the bearings are set up in a face-to-face
configuration. In Figure 4.20 the gear is overhung mounted. A pair of tapered
roller bearings takes thrust in one direction while a single bearing reacts thrust
in the opposite direction. The bearings are set up in a back-to-back configura­
tion. Care must be taken in the assembly to provide sufficient endplay for
thermal expansion of the shaft. In calculating thrust loads on a shaft supported
by tapered roller bearings it must be remembered that due to the angle of
contact, a radial load on one bearing induces an axial load that must be reacted
either by an external load or the other bearing. The induced axial load depends
on the bearing internal geometry, and methods of calculation are included in
catalogs of tapered roller bearings.
176 Gear Drive Systems

Bearing Lubrication

The type of lubrication to which rolling element bearings are exposed in gear
units is dictated by the gear oil requirements. Gear boxes discussed in this book
typically incorporate oil splash or jet lubrication systems and the bearing lubri­
cation discussion will be limited to these areas. The types of oil used are
discussed in Chapter 5.
The functions of a rolling element bearing lubricant are:
1. Provide an oil film between contacting elements.
2. Provide a cooling medium.
3. Protect bearing surfaces against corrosion.
Very little oil is required to maintain a satisfactory oil film; however, for high­
speed highly loaded bearings significant oil flow must be supplied to perform the
cooling function.
In a splash lubrication system the gearbox is filled with oil to a level that
the gears and bearings are dipping into the lubricant as they rotate. At low
speeds the heat generated can be dissipated through the casing and the unit
reaches a satisfactory equilibrium temperature. Typical of such applications are
electric or hydraulic motor speed reduction units. If heat generation becomes
excessive the casing can be air or water cooled and an integral splash lubrica­
tion system may still be satisfactory. As speeds and loads increase a circulating
system may be necessary with an external oil cooler.
In high-speed boxes, gear pitch line velocities of 5000 fpm or more, jet
lubrication is used. At high speeds it is impractical to allow the components to
rotate through an oil bath since the churning would create excessive heat and
Bearings and Seals 177

power loss. In some high-speed, light-load applications oil mist lubrication can be
used to avoid churning losses in the bearing.
With jet lubrication one or more jet streams are directed on the bearing at
the gap between the retainer and race. It is good practice to use more than one
jet to ensure against oil starvation if a jet clogs. The jet diameter should be a
minimum of 0.030 in. also to avoid clogging.
Flow through an oil jet can be calculated as follows:

where
Q = oil flow, in.3/sec
A = jet area, in.2
K = discharge coefficient (assume 0.65)
g = acceleration due to gravity, 386 in./sec2
Ap = pressure drop across orifice, lb/in2
w = specific weight, lb/in.3
In order to calculate the amount of flow required to cool the bearing an estimate
of the heat generated is needed. Through experience, empirical methods have
been generated which approximate the losses due to friction and oil churning.
Two such methods will be presented. The first method [6] estimates losses as
T = fRW
where
T = frictional torque, in.-lb
f = friction coefficient
R = shaft radius, in.
W = load, lb
The following table lists friction coefficients for various bearing configurations.

Bearing type Friction coefficient

Radial ball bearings 0.0015
Self-aligning ball bearings 0.0010
Angular-contact bearings 0.0013
Pure thrust ball bearings 0.0013
Cylindrical roller bearings 0.0011
Spherical roller bearings 0.0018
178 Gear Drive Systems

From the friction torque the horsepower or heat generation can be calculated:

63,025
where N is the bearing rpm.
Q = HP(42.44)
where Q is the heat generated in Btu/min. This method does not consider the
amount of oil flow or the oil viscosity directly. Another empirical procedure
that includes these parameters is [7]:

Q = B[(DN)1-5W°-07M°-4 V - 25]
where
Q = heat generated, Btu/min
D = bearing bore, mm
N = bearing rpm
W = bearing load, lb
(i = dynamic viscosity, Reyns (lb-sec/in.2)
M = oil flow, lb/hr
The following table lists values for the coefficient B:

Type of bearing B coefficient

Angular contact 10.1 X i o - 7
Radial ball 4.46 X 10~7
Cylindrical roller 6.46 X IO"7

For a given flow and heat loss the oil temperature rise across the bearing can be
calculated:

AT = Q/M(cp)
60
where
AT = temperature rise, °F
M = oil flow, lb/hr
Cp = specific heat, approximately 0.5 Btu/lb-°F
The conversion from GPM to lb/hr for oil is approximately
1 GPM ss 450 lb/hr
Bearings and Seals 179

Figure 4.21 Progression of fatigue spalling. A. Incipient fatigue spalling. B. More

advanced spalling. C. Greatly advanced spalling. (Courtesy of SKF Industries, Inc.,
King of Prussia, Pa.)
180 Gear Drive Systems

Bearing Failures

Rolling element bearing failures are characterized by one or more of the follow­
ing operating conditions:
1. Excessive vibration
2. Excessive noise
3. Overheating
4. Chip generation
5. Hard turning or excessively loose shafts
These problems can be a result of a great variety of errors but can usually be
categorized as either load or lubrication related.
Lubrication-related failures occur when the supply of lubricant is insuffi­
cient or misdirected. The mating bearing surfaces, without a proper oil film,
come into intimate contact, resulting in wear, heat generation, and thermal
expansion. The thermal expansion reduces the bearing internal clearance and
therefore increases the bearing load and the failure becomes self-perpetuating.
Quite often a failure that appears to be due to overload masks a lubrication
problem that led to loss of internal clearance and internally generated loading.
Load-related failures can be a result not only of the operating loads but
also of forces applied during assembly. Also, the effect of operating loads can
be magnified due to defective bearing seats on shafts or in housings or
misalignments.

B
Figure 4.22 Normal radial load-bearing patterns. (Courtesy of SKF Industries,
Inc., King of Prussia, Pa.)
Bearings and Seals 181

The mode of failure related to load that is most common is fatigue spall­
ing. Figure 4.21 shows the progression of fatigue spalling. The small spall in
Figure 4.21A began with a crack which probably originated below the race
surface. This initial spall will advance to the size shown in Figure 4.2IB, at
which point the vibration and noise level of the unit will have increased signifi­
cantly. Left unattended, the spalling will progress in proportion to speed and
load as shown in Figure 4.21C.
Sometimes a clue to the cause of bearing problems can be found by
examination of the pattern of the load path on the bearing races. Figure 4.22
shows the normal load pattern of a radial bearing. Figure 4.22A illustrates a
rotating inner ring operating with a stationary load. The pattern of a rotating
outer ring with a stationary load or a rotating inner ring with a load rotating in
phase is shown in Figure 4.22B. Figure 4.23 illustrates load patterns on a thrust
bearing. Figure 4.23A is a nornal pattern which stays within the raceway, not
reaching out to the edge. In Figure 4.23B the thrust load is excessive and the
balls contact the race edge, resulting in a load concentration. Figure 4.23C
illustrates a normal contact pattern for a bearing experiencing combined radial
and thrust load. Figure 4.24 shows load patterns that reflect problems. Figure
4.24A illustrates a bearing that is internally preloaded. This may be due to loss
of clearance because of excessive press fits on the shaft or in the housing or
possibly thermal expansion. Figure 4.24B shows the load pattern produced by
an out-of-round housing pinching the bearing outer ring. Figure 4.24C shows the
load zone when the outer ring is misaligned relative to the shaft, and Figure
4.24D the pattern when the inner ring is misaligned with respect to the housing.
Two types of pitting failures that occasionally occur and are hard to
diagnose are false brinelling and electrical pitting. False brinelling is a condition
caused when the gear unit is subject to vibration while the shafts are not rotat­
ing, such as in transit during shipping. It is usually characterized by polished
depressions spaced equal to the distance between rolling elements. False
brinelling is a form of fretting corrosion.
Electrical pitting occurs when a current seeks ground by passing through a
bearing. The current is broken where the balls or rollers contact the raceway
and arcing results, causing high temperatures and pitting damage.

Bearing Costs

Several variables influence the cost of gearbox bearings:

1. Obviously, when buying bearings in production quantities rather than proto­
type, lower prices prevail. The quantity can also determine whether the
bearings are bought from a manufacturer, distributor, or supply house,
which can affect the price significantly.
182 Gear Drive Systems

B
Figure 4.23 Normal thrust load-bearing patterns. (Courtesy of SKF Industries,
Inc., King of Prussia, Pa.)
Bearings and Seals 183

Figure 4.24 Abnormal bearing load zone patterns. (Courtesy of SKF Industries,
Inc., King of Prussia, Pa.)
184 Gear Drive Systems

Table 4.1 Cost Comparison, 75-mm Bearing

Basic Cost of
load Limiting basic load
rating speed Relative rating
Bearing type (lb) (rpm) cost (lb/dollar)
Self-aligning ball 13,700 4,200 1.69 810
bearing
Single-row, deep 19,600 4,200 1.32 1.480
groove ball bearing
Angular contact (40°) 21,500 4,200 2.10 1,030
ball bearing
Maximum capacity, 25,700 4,200 1.52 1,700
deep groove ball
bearing
Cylindrical roller 36,500 4,200 3.53 1,030
bearing
Double row, deep 30,200 2,900 2.87 1,050
groove ball bearing
Tapered roller 50,100 2,460 1.00 5,010
bearinga
Spherical roller 68,000 2,500 4.00 1,700
bearing
aBore: 3.000 in., O.D.: 5.909 in.
Source: Ref. 8.

2. Design options such as special material or processing add greatly to the basic
cost.
3. Higher-than-standard precision requirements or internal clearances which are
nonstandard have a significant price effect.
An interesting study was conducted investigating the comparative cost of
bearings on a basic load rating per dollar basis [8]. Various configurations of the
same two sizes, 55- and 75-mm bore, were evaluated to determine how much
load capacity each bearing offers per dollar of cost. From Tables 4.1 and 4.2 it is
seen that tapered roller bearings offer the best value on this basis. Of course,
many applications have design conditions which dictate that other types of
bearings are required.
It must be remembered that the initial bearing cost is not the only expen­
diture to be considered. If failure occurs, the expense of disassembly and
reassembly must be considered, and also the downtime cost.
Bearings and Seals 185

Table 4.2 Cost Comparison, 55-mm Bearing

Basic Cost of
load Limiting basic load
rating speed Relative rating
Bearing type (lb) (rpm) cost (lb/dollar)
Self-aligning ball 4,630 6,500 1.99 982
bearing
Single-row, deep 7,500 6,500 1.42 2,220
groove ball bearing
Angular contact (40°) 8,010 6,500 2.39 1,411
ball bearing
Maximum capacity, 9,830 6,500 1.70 2,430
deep groove ball
bearing
Cylindrical roller 10,300 6,500 5.70 762
bearing
Double row, deep 11,400 4,500 2.92 1,650
groove ball bearing
Tapered roller 15,000 3,720 1.00 6,330
bearing21
Spherical roller 19,300 4,000 6.02 1,350
bearing
aBore: 2.1653 in.
Source: Ref. 8.

Journal Bearings

A simple radial bearing is shown in Figure 4.25. At rest, the shaft (journal) will
lay on the bottom of the bearing with a load W, the shaft weight. At this point
the lubricant is squeezed out and metal-to-metal contact occurs (Figure 4.26A).
As the shaft begins to rotate (Figure 4.26B) it will climb up the bearing wall
and be slightly off-center. The rotation of the shaft will tend to pull oil into
the interface between the shaft and bearing, a wedge-shaped zone. The inlet to
this converging region would like to take more oil in than the outlet will allow
out. The jamming of the fluid into the converging region creates a pressure
distribution between the journal and bearing as shown in Figure 4.27. The high
pressure in the center forces the fluid to slow down at the inlet and speed up at
the outlet, so that the flow coming in equals the flow going out. The pressure
generated creates a lifting force which separates the journal and bearing with an
oil film (Figure 4.26C). This pressure is the basis of hydrodynamic lubrication
and the load capacity of the bearing depends on the hydrodynamic pressure
186 Gear Drive Systems

OIL INLET

Figure 4.26 Journal bearing operation.

Figure 4.27 Hydrodynamic pressure distribution.

Bearings and Seals 187

Figure 4.28 Regimes of lubrication.

generated. Increasing the fluid viscosity or the shaft velocity will increase the
hydrodynamic pressure and therefore increase the film thickness. Typical hydro-
dynamic bearings operate with film thicknesses in the range 0.0005 to 0.002 in.
As a bearing accelerates from rest it passes through three regimes of
lubrication. The first regime is boundary lubrication, where there is metal-to-
metal contact. The second is mixed lubrication, where there is a transition from
boundary, to hydrodynamic lubrication, where the bearing operates with a full
fluid film. These regimes are illustrated in Figure 4.28, where the coefficient of
friction is plotted against a bearing parameter:

Bearing parameter = ^

where
Z = viscosity, cP
N = shaft rpm
P = unit loading, psi
The unit loading is defined as
188 Gear Drive Systems

where
W = radial load, lb
L = bearing length, in.
D = bearing diameter, in.
The minimum ZN/P value to assure full film lubrication varies with the applica­
tion. As a rule of thumb, for lightly loaded high-speed bearings with low-
viscosity oils it should exceed 150. Low-speed, high-load applications with heavy
oils may have ZN/P values as low as 10.

Pressure-Fed Bearings

Gearbox journal bearings are usually supplied oil under pressure. This assures a
reliable source of lubricant and also provides a cooling medium. Feed pressures
are typically in the range 20 to 50 psig. The oil flow to the bearing must be
sufficient to replace oil lost by leakage at the bearing ends. Usually, the feed oil
passages are sized sufficiently large such that the flow is metered by the bearing
itself. There are various designs of pressure-fed radial bearings in use, the dif­
ference being in the shape and location of the oil feed grooves and the shape of
the bearing bore.
Oil distribution grooves should not be located in the load zone since this
will disrupt the hydrodynamic pressure distribution and reduce the load-carrying
capacity.
A problem that arises with high speed bearings when they are lightly
loaded is oil whip or whirl. It is a vibration phenomenon which occurs at either
half the rotational speed or the natural frequency of the shaft when the shaft is
operating well above the natural frequency. One explanation of whirl is that a
wedge of oil is traveling around the bearing at the average oil velocity (one-half
shaft velocity) or at the shaft critical velocity. Bearing bore designs other than
the simple full round bearing (Figure 4.25) are meant to minimize this whirl
tendency.
Following are descriptions of some commonly used gearbox radial
bearings:
1. Axial groove (Figure 4.29A). This type of bearing has a cylindrical bore and
one or more axial oil spreader grooves extending almost the full length of the
bearing. It has high load capacity but at light loads, usually under 150 psi, is
quite susceptible to oil whip.
2. Elliptical (Figure 4.29B). The elliptical bearing has a large clearance in one
direction and a smaller clearance 90° away. It is manufactured by placing
shims at the split between two halves of a cylindrical bearing and then
machining the bore. The shims are removed, producing a variation in clear­
ance around the bore, which is not truly elliptical. This type of bearing
Bearings and Seals 189
X<*-1

< ..........0 ................>

< ....... ° . >

Figure 4.29 Types of pressure-fed radial bearings.

allows larger oil flows than a cylindrical axial groove design and therefore
affords more cooling. It is also somewhat more stable as regards oil whip.
3. Offset half (Figure 4.29C). This is a modification of an axial groove bearing
where the upper and lower halves are offset. It has proved successful in elim­
inating oil whip. A disadvantage of this design is that it can accommodate
shaft rotation in only one direction.
190 Gear Drive Systems

4. Tilting pad (Figure 4.29D). The individual pads of this design are free to
pivot and follow shaft excursions; therefore, forces produced in the bearing
are incapable of driving the shaft into an unstable mode and operation is
smooth. Each pad tilts such that a wedge-shaped oil film is formed which
tends to center the shaft. The pivot can be located anywhere from the
center of the pad to near the trailing edge. In one design the pads are pivoted
near the trailing edge and the other edge is forced toward the shaft by
springs. This preload adds significantly to high-speed stability. A disadvan­
tage to this type of bearing is that it must be flooded with oil; therefore,
required flows are high with accompanying power losses. Also, the bearing
is mechanically complex and the parts are prone to fretting.

Relative Bearing Costs

Of the pressure-fed bearing designs the axial groove configuration is easiest to

fabricate and has the lowest cost. Elliptical bore and offset half bearings are
somewhat more expensive and tilting pad bearings are significantly more
expensive.

Design and Rating of Radial Journal Bearings

The gearbox designer tends to be concerned with simple criteria such as the bear­
ing unit loading, its length-to-diameter ratio, and the clearance required when
defining radial journal bearings. These concerns are not sufficient to adequately
design a bearing and the bearing specialists are turned to for optimization of the
final design. The significant parameters to be analyzed are minimum film thick­
ness, maximum bearing surface temperature, and maximum hydrodynamic
pressure. These parameters must be traded off against one another to arrive at
the optimum design. The mathematical design of journal bearings is beyond the
scope of this book, but the more commonly used parameters will be discussed:
Unit Loading Typical gearbox radial bearings can operate successfully to
unit loadings of approximately 550 psi. By unit load is meant the radial load
divided by the projected bearing area:

LD
where
L = bearing length, in.
D = bearing diameter, in.
More stringent criteria are used in some cases where extreme reliability is
desired. For instance, some company specifications limit unit loading to 350 psi.
Bearings and Seals 191

Of course, the loading allowed is dependent on the bearing material. This is dis­
cussed in a later section.
Clearance Many designers as a rule of thumb use 0.001 in. of diametral
clearance per inch of shaft diameter. In other words, a 6-in.-diameter journal
would operate in a 6.006-in.-diameter bearing. This rule of thumb does not take
shaft speed into account and a better estimate of minimum clearance is [9].
N°-25D 10 3
Minimum diametral clearance = --------- +— X 10
6 D
where
N = rpm
D = shaft diameter, in.
When a bearing is pressed into a housing the bore will tend to collapse and thus
affect the clearance. Clearance, therefore, should be specified after assembly or
the bore made sufficiently oversize to accommodate the collapse after press fit,
which may be as much as 75% of the press.
Length-to-Diameter Ratio The length-to-diameter ratio of gearbox
journal bearings is usually around 1.0, although it can vary from 0.3 to 2.0. In
shorter bearings there is a reduction in load-carrying capacity due to excessive
end leakage. Longer bearings are susceptible to end loading due to misalignment.

Lubrication

The amount of oil leakage from both ends of a journal bearing can be estimated
as follows [10]:
rc3ps
Q = 0.816
ML
where
Q = total flow, gal/min
r = shaft radius, in.
11 - oil viscosity, Reyns
ps = oil supply pressure, psi
L = uninterrupted bearing land length, in.
c = radial clearance, in.
The viscosity ju should be taken at the bearing operating temperature, which can
be estimated as the average of oil temperature into and out of the gearbox. If the
bearing has a full circular oil feed groove, the length L will be approximately half
of the overall bearing length.
192 Gear Drive Systems

■A
v i e w AA

Figure 4.30 Simple thrust washer.

The oil chosen will depend on the unit speed and loading conditions. High­
speed gear units typically use light turbine oils of approximately 150 SUS
viscosity at 100°F. Lower-speed, highly loaded units will use heavier oils and
occasionally EP additives.
After adequate oil flow, oil cleanliness is a prime factor controlling bearing
life. The need for cleanliness is obvious when one considers that the magnitude
of film thickness is sometimes as low as 0.0005 in. Proper cleaning of the gear
unit prior to operation and continuing filtration of the lubricant cannot be too
highly stressed.

Thrust Bearings

Thrust Washers Figure 4.30 illustrates the simplest type of journal thrust
bearing, a flat stationary surface which reacts thrust from a flat runner which is
either integral or attached to the rotating shaft. Because there is no taper to
promote hydrodynamic action, this type of bearing normally should not be
designed for unit loadings above 75 psi. Higher loads are possible but require
detailed design and development.
The bearing is lubricated by directing oil to the inside diameter, the oil
passing through radial oil feed grooves to the outside. To simplify manufacture
an even number of grooves is usually specified. Various groove designs are used,
which are usually as deep as they are wide. The load-carrying lands should be
approximately square.
Tapered Land A tapered land thrust bearing is illustrated in Figure 4.31.
The pad lands between oil grooves are tapered such that the runner will carry
oil into a wedge-shaped region and build up a load-carrying hydrodynamic
pressure. The taper may be simple or compounded, being larger at the inside
Bearings and Seals 193

diameter of the pad than at the outside diameter. This is done in an attempt to
equalize oil flow across the pad and maintain an even temperature distribution
to reduce distortion. To maintain some load-carrying ability when the bearing is
starting up, the taper only extends for approximately 80% of the pad. The
width-to-length ratio of the pads is approximately 1.0. Oil is fed to the inside
diameter of the bearing and distributed outward through radial oil grooves. The
grooves are usually dammed at the ends with a small chamfer cut to pass a

Figure 4.32 Tilting pad thrust bearing. (Courtesy of Glacier Metal Company
Ltd., Wembley, England.)
 feet/ sec
ABSORBED-kW
POWER

Figure 4.33 Tilting pad thrust bearing losses. (Courtesy of Glacier Metal
Company Ltd., Wembley, England.)

194
Bearings and Seals 195

limited amount of oil and prevent the accumulation of dirt at the groove ends.
The amount of taper varies with the bearing size. A 1-in.2 pad may have a
0.004.in. taper, while a 7-in.2 pad may have a 0.008-in. taper.
Tapered land thrust bearings are used at unit loadings up to approximately
550 psi. In some critical applications where extreme reliability is required, they
are derated to as low as 150 psi. Disadvantages of this type of bearing are its
inability to accommodate significant misalignment and that because of the taper
it has high load-carrying capacity in one direction of rotation only.
Tilting Pad This type of bearing, sometimes called pivoted shoe or Kings­
bury, consists of a number of pads which are independently free to pivot in
order to provide a tapered oil film (Figure 4.32). At rest the pads are parallel to
the shaft face. As the shaft rotates, an oil film is generated and each pad tilts to
an angle such that the oil pressure is evenly distributed. The oil is usually fed
into the center of the bearing near the shaft and exits at the bearing outside
diameter. As shown in Figure 4.33, the bearing area is sealed to ensure that the
bearing operates in a flooded condition. Because of this, tilted pad bearings tend
to generate more heat than other types and Figure 4.33 illustrates the losses in
the bearing due to shear of the oil film and also due to churning. The churning
losses can be significantly reduced by directing the lubrication to the pad face
only (Figure 4.34).
Tilting pad bearings can accommodate shaft rotation in both directions
if the pad pivots are located in the middle of the pads. To obtain even greater
tolerance for misalignment pads are sometimes mounted on leveling plates,
Figure 4.35.
Tilting pad bearings can withstand unit loadings of 550 psi or more and
are often specified in applications where long life and reliability are paramount.
The disadvantages of this type of bearing are high cost, oil flow, and power loss.
Bearing Materials Because a journal bearing experiences a variety of lubri­
cation regimes from startup to full load and speed, the material must have a
blend of characteristics to accommodate all the demands placed on it. This
always requires a compromise between a high strength and temperature capacity
and good surface characteristics, such as resistance to seizure and compliance.
A discussion of the most important bearing material characteristics follows:
1. Seizure resistance At startup and on occasion during operation the oil film
breaks down and metal-to-metal contact between the journal and bearing
will occur. The bearing material must be capable of withstanding this
contact without welding, scoring, or tearing.
2. Ability to absorb foreign objects If dirt particles enter the clearance space,
it is desirable for the particles to become embedded in the bearing material,
where they cannot score or wear the shaft.
196
OIL PAD SPACIN G STO P
DISCHARGE

Oil flow in thrust bearing

with directed lubrication

Figure 4.34 Directed lubrication of tilting pad thrust bearing. (Courtesy of Glacier Metal Company Ltd., Wembley,
England.)
Bearings and Seals 197

Figure 4.35 Leveling plate design for greater accommodation of misalignment.

3. Compliability The bearing material modulus of elasticity should be low

such that if the shaft tends to touch the bearing ends due to misalignment
the bearing will conform locally and the deformation will prevent severe
rubbing and excessive temperature rise.
4. Fatigue Resistance The bearing must have the ability to withstand repeated
stress cycles. In the case of gearbox bearings the load is generally steady;
however, unbalance can lead to a rotating load with fatigue potential.
5. Corrosion Resistance If the lubricating oil contains acid or becomes acidic
due to oxidation, the bearing material must be capable of withstanding the
acidic attack.
The requirements of seizure resistance, embeddability, and compliability lead
toward a soft bearing material such as babbitt, while fatigue strength requires a
harder material. In the majority of gearbox bearings the bearing construction
consists of a thin surface layer of babbitt bonded to a mild steel backing. In
some cases a three-layer or trimetal construction is used with a higher-strength
bearing material bonded to a steel backing and overlaid with a thin babbitt. A
discussion of the most widely used materials follows:
1. Babbitt. Babbitts have the best seizure, embeddability, and compli­
ability characteristics of all bearing materials. Two types are in common use;
tin base and lead base. Table 4.3 lists the composition of these materials. The
tin-base material is slightly more desirable in terms of seizure and corrosion
resistance; however, the lead base is less expensive. A lead-base babbitt with a
small percentage of tin is used in many applications as an economic compromise.
Babbitts are inherently weak and also temperature limited. They should
not be used at operating temperatures over 250°F. Babbitt bearings in gearboxes
are usually bonded to a mild steel backing material and the babbitt thickness is
on the order of 0.030 to 0.060 in. The strength of a babbitt surface decreases
with its thickness and in very high load applications such as crankshafts, which
198
Table 4.3 Composition of Bearing Materials
Nominal composition (%)
Material Specification Copper Tin Lead Zinc Antimony Nickel Arsenic
Lead-base babbitt SAE 14 V2 10 74 15 J/2
Tin-base babbitt SAE 12 3 J/2 89 IV2
Copper-lead SAE 480 65 35
Leaded bronze SAE 792 80 10 10
Leaded bronze SAE 794 72 3 23 2
Aluminum 1 1 6 1
Bearings and Seals 199

are prone to fatigue, babbitt thickness may be as low as 0.001 in.; however, in
gearboxes conformability and embeddability are important and thicker layers
are used.
2. Copper-Lead. This is the simplest of the copper-based materials and
contains from 20 to 40% lead. The material has excellent fatigue strength and
is capable of carrying heavy loads at high temperatures. Compared to babbitt,
however, copper lead has poor seizure resistance, embeddability, and compli­
ability. The surface behavior can be improved by using hardened journals.
3. Bronze. There are three categories of bearing bronze; lead bronze,
tin bronze, and high-strength bronze. High-strength bronze may contain alu­
minum, iron, manganese, silicon, nickel, or zinc in varying percentages. Strength
improves from lead to tin to high-strength bronze, and shaft compatibility
characteristics deteriorate. Bronzes are cost effective since they can be easily
cast and machined and do not require a steel backing.
4. Aluminum. Aluminum has good fatigue resistance and corrosion
resistance but does not have the surface properties of babbitt. Hardened shafts
should be used in conjunction with aluminum bearings.
5. Trimetal composite bearings. In order to take advantage of the
surface characteristics of babbitt yet improve bearing strength and temperature
resistance, trimetal bearings have come into use. An intermediate layer of
approximately 0.020 in. of copper-lead or leaded bronze is sandwiched between
a mild steel backing and an overlay of babbitt approximately 0.002 in. thick.

Shaft Definition

Many journal and thrust bearing problems are a result of shaft discrepancies such
as improper geometry or poor surface finish. The important geometric charac­
teristics are taper, out of roundness, and grinding discrepancies such as waviness,
chatter, or lobing, which must be closely controlled.
Shaft taper in the bearing area should be limited to 0.0002 in. per inch of
length. Out of roundness should be limited to 0.0005 in. for shaft diameters up
to 5 in. and 0.001 in. for shafts above 5 in. in diameter. Shaft surface finish
should be a maximum of 20 rms with a preferred finish of approximately 10 rms.
Journals operating with babbitted surfaces can be made of soft steel with
a minimum hardness of 200 bhn. Aluminum and copper-lead materials can be
operated with shafts of 300 bhn minimum hardness. Where loads are high a
journal hardness of Rc50 is recommended. In general, it is desirable to harden
shaft surfaces in any case, in order to obtain longer life and improved wear and
abrasion characteristics. In gearboxes the bearing journals quite often are integral
with a pinion and therefore made of the gear material and hardened to the gear
tooth specification and ground. In the case where gears are assembled on a shaft,
the shaft might be SAE 4140 or 4340 steel and the journals nitrided and ground.
 200
Gear Drive Systems

W H k
■MMNP

, |jV.
^11 M
I '

K S ! ; ! £ 7 ,tata(c“«^»'«“»
Bearings and Seals 201

Journal Bearing Failures

Unlike rolling element bearings, fatigue failures are not common in gearbox
journal bearings. Failures that occur are more likely to be associated with con­
tamination of the lubricant, insufficient lubrication, dynamic excursion of the
shaft, or faulty assembly.
Lubricant contamination. Abrasive materials in the lubricant may be a
result of insufficient cleaning of the machined components at assembly, dirt
entering the unit through breathers or bypassed oil filters, or wear particles
generated inside the unit. Figure 4.36A shows a babbitted surface which has
been scored and pitted by dirt, and Figure 4.36B shows surface distress to a
greater extent caused by dirt particles. Figure 4.36C illustrates a thrust pad
scored circumferentially by dirt particles.
The bearing in Figure 4.36A can be cleaned and reused provided that the
wear experienced is not excessive and the bearing clearance remains within
specification. Of course, the source of contamination must be eliminated and
the lubrication system flushed out before continuing operation.
Lubricant cleanliness cannot be too greatly emphasized when operating
journal bearings. From the assembly area, where all oil passages should be
mechanically cleaned to eliminate machining chips, to the test stand, where the
unit should be extensively flushed, and in the field, where oil must be carefully
filtered, continuous care must be taken not to contaminate the oil. Filtration
should be a maximum of 40 jum, and lower filtration levels to 10 pm are
beneficial. When servicing the unit in the field care must be taken not to intro­
duce dirt through filler ports or inspection covers.
Wiping o f Bearing Surfaces. When the rotating journal and the bearing
metal touch during operation the rubbing causes melting and smearing, as shown
in Figure 4.37. Wiping may be due to insufficient clearance, overheating which
closes down the clearance,high transient loads, shaft vibration due to unbalance,
or dynamic journal instabilities. Figure 4.37A shows both halves of a wiped
babbitted bearing and Figure 4.37B illustrates an overlay-plated copper-lead
bearing wiped on half the circumference. When wiping occurs if the surface
distress is light, the bearing can be cleaned and reused provided that the wear
does not result in excessive clearance. The cause of the wipe should be identi­
fied and corrected. For instance, if a vibration survey shows synchronous
vibration, the shaft should be balanced. If oil whip is detected, a different profile
bore is indicated.
Corrosion. Corrosion of a bearing is a result of chemical attack by
reactive materials in the oil stream. The most common problem is oxidation
products formed in the oil, which corrode materials such as lead, copper,
cadmium, and zinc. Figure 4.38 shows the severely corroded surface of a
202 Gear Drive Systems

Figure 4.37 Wiping of bearing surface. (Courtesy of Glacier Metal Company

Ltd., Wembley, England.)

copper-lead bearing. Lead reacts rapidly with oil oxidation products. To solve
this problem oxidation inhibitors are incorporated in the lubricant and pro­
tective alloying elements such as tin are included in the lead babbitt formulation.
Faulty Assembly. The fit of the bearing in the housing is a potential
source of difficulty. If the bearing is too loose in the housing fretting corrosion
can result. This also can occur if the bearing housing design is too flimsy. Too
tight a press fit can cause bore distortion. Another cause of bore distortion is
entrapment of foreign particles between the bearing and its housing during
Bearings and Seals 203

m -" V / . " i

• Kk ^ y ij
/ OH» V. j r- .■

jv.rfrov5*
.v;v ^ .«.--* ...> v j
•'

V ...........1

Figure 4.38 Bearing corrosion. (Courtesy of Glacier Metal Company Ltd.,

Wembley, England.)

assembly. This will result in a localized bore distortion and potential overheat­
ing. Misalignment of the shaft with respect to the housing will lead to uneven
bearing wear, as shown in Figure 4.39.

Salvage Procedure

Following a journal bearing seizure the shaft may be scored or worn to such an
extent that rework is required. There are two potential rework procedures. One
is to grind the shaft down and press fit on a sleeve, which is ground to size after
assembly. The sleeve must be sufficiently thick to withstand the press fit stresses
and the fit should be approximately 0.001 in. per inch of shaft diameter.
Another rework method is to grind the shaft down at the distressed area and
build it up with chrome plate or a metal spray. Following is a rework procedure
for hard chrome plating:
1. Grind the diamter 0.015 in. under the low limit.
2. Bake after grinding for 4 hr at 275°F.
3. Shot peen the rework area using 170 shot to an intensity of 0.012 to 0.014A.
4. Hard chrome plate the rework area and bake after plating for 4 hr at 275°F.
Plate to 0.004 in. above the high limit.
5. Grind to finished size.
Care must be taken in grinding the chrome plate since abusive grinding can over­
heat the base metal, resulting in temper and cracks. These cracks are masked by
the chrome plate and cannot be detected by magnaflux inspection. The cracks
can propogate through the chrome plate during operation, lifting off the plating
and causing bearing failure.
204 Gear Drive Systems

Figure 4.39 Bearing pattern due to misalignment. (Courtesy of Glacier Metal

Company Ltd., Wembley, England.)

SEALS

Gearbox shaft seals are usually either elastomeric lip seals or noncontacting
labyrinth seals. Elastomeric lip seals are generally found in low-speed applica­
tions in conjunction with rolling element bearings. The limiting surface speed of
a lip seal is on the order of 3600 fpm. Somewhat higher speeds are possible but
require careful design and development. When journal bearings support a shaft,
the shaft movement may be excessive for successful lip seal operation.
Labyrinth seals because they are noncontacting are not speed limited and
also do not introduce any friction torque into the system. They are used in
critical high-speed applications because there is little likelihood of failure. On
occasion, carbon face seals are incorporated in gear applications where speeds are
high and a more positive seal than a labyrinth is desired.
Labyrinth and lip seals will allow oil leakage if the gearbox internal
pressure is higher than the external ambient pressure. This situation may occur
for the following reasons:
1. The gearbox may be vented to an area where the pressure is higher than
ambient pressure. This causes the unit to be back-pressured, resulting in oil
leakage.
Bearings and Seals 205

2. On occasion the gearbox may serve as the sump for the scavenge oil of
another piece of equipment, such as a turbine. This oil may be aerated and
at a pressure higher than ambient.
3. The area external to the gearbox may be at a slight vacuum and the gear
unit, if vented to atmosphere, will be at a higher internal pressure.
4. If the gear unit is not vented, heat generated in the box can expand the air,
creating a pressure differential with the outside.
If any of the conditions noted above result in oil leakage, several solutions are
possible:
1. Lip seals with greater interference or higher spring force can be used; how­
ever, this may result in excessive wear.
2. Labyrinth seals can be internally pressurized or internally drained to prevent
oil leakage.
3. A face seal can be incorporated which can seal against small pressure
differentials.
The following paragraphs will describe the seal configurations most commonly
used in gearboxes.

Labyrinth Seals

Labyrinth seals are devices that limit leakage between a rotating shaft and a sta­
tionary housing by maintaining a close radial clearance between the two. Figure
4.40 illustrates a typical labyrinth seal where several stages of knives deter oil
leakage from the gearbox cavity. The labyrinth pictured has two sets of knives
with an interstage area which can be used for one of two purposes. A drain can
be located between the knives such that any oil leakage past the first stages will
be diverted back into the gearbox.
The interstage area can also be used to introduce air pressurization and
maintain the air side of the seal at a higher pressure than the oil side. Usually, a
pressure differential of 1 or 2 psi is sufficient to effectively eliminate oil leakage.
This technique is used when a source of bleed air such as a turbine is readily
available. An estimate of the airflow through a labyrinth seal to create a desired
pressure differential can be made with the following equation [11]:
M = K xK 2 K 3 K 4P0A

where

M = airflow, lb/sec
K2 = coefficient dependent on the ratio of radial gap G to knife tip width
W (Figure 4.41)
206 Gear Drive Systems

K3 = coefficient dependent on the ratio of radial gap G to knife pitch P

and number of knives (Figure 4.41)
Ki = coefficient dependent on pressurization air temperature (Figure
4.41)
K4 = coefficient dependent on the pressure ratio (low pressure/high
pressure) across the seal and the number of knives (Figure 4.41)
P0 = high pressure, psia
A = gap area, in.2
As an illustration of this analysis, assume that a pressure of 2 psig (16.7 psia) is
desired in the interstage area of the seal shown in Figure 4.40. The air is at a
temperature of 100°F.
14 7
Pressure ratio = ----- = 0.88
16.7
Bearings and Seals 207

2 .o f NO. OF KNIVES

0 0.02 0.04 0.06 0.08

GIP

Figure 4.41 Labyrinth seal air leakage. (From Ref. 11.)

208 Gear Drive Systems

p = 0.10
w = 0.006

Number of knives = 4
= 0.0325
k2 = 1.0
k3 = 1.5
k4 = 0.2

A = 0.104

Po = 16.7

M = 0.017 lb/sec airflow

If volume flow is desired, the conversion from pounds per second to standard
cubic inches per minute (SCIM) is
5000 SCIM ~ 0.004 lb/sec
Therefore,
M = 21250 SCIM = 12.3 SCFM (airflow through one-half the seal)
In the interest of limiting leakage it would be advisable to reduce the clearance
to the smallest possible amount. The practical clearance limit must take into
account bearing tolerances, dynamic deflections, and thermal distortions. If the
shaft motion exceeds the clearance, a rub will occur which will wear or score the
knives and the shaft. In journal bearing gear boxes the radial clearance is in the
order of 0.010 in. To achieve closer clearances it is possible to coat the knives
with an abradable material such as tin babbitt. If the shaft touches, no harm will
result and minimum gaps can be maintained.

Elastomeric Lip Seals

Figure 4.42 illustrates a typical elastomeric lip seal. By elastomeric it is meant

that the sealing lip is made of a synthetic rubber compound. Oil sealing is accom­
plished through an interference fit between the flexible sealing lip and the shaft.
Because the elastomer can lose tension and elasticity during operation, a garter
spring is usually incorporated on the lip to ensure pressure at the sealing
interface. In the design of Figure 4.42 a dirt excluder lip on the air side is used
to shield the oil sealing lip from contaminants. Because the dirt exclusion lip is
unlubricated its design interference will be less than the oil seal lip interference.
Bearings and Seals 209

The amount of interference and lip contact pressure required depends on the
following variables:
Elastomer compression set
Thermal expansion
Shaft eccentricity
Shaft-to-bore misalignment
Typical lip-to-shaft interferences, or pinches as they are sometimes called, are on
the order of 0.035 in.
A paradox of lip seals is the fact that although they are sealing devices, in
order to operate properly there must be a lubricant film between the lip and the
shaft. Successful sealing and long seal life depend on maintenance of this film,
which is usually about 0.0001 in. thick. If no film is present, the seal lip will
wear out or tear. The majority of lip seal applications, about 80% will leak
slightly; on the order of 0.002 g/hr or about 1 drop every 11 hr [12]. About
15% will leak 0.002 to 0.1 g/hr. This level of leakage may be unacceptable.

Shaft Specification

Medium carbon steel such as SAE 1035 or 1045 or stainless steel shafts are the
best sealing surfaces, but chrome or nickel plated shafts are also acceptable. The
shaft hardness should be Rc 30 minimum. Shaft finish should be 10 to 20 juin.
and the shaft should be plunge-ground with no machine lead. Wear sleeves are
commercially available which can be pressed over the shaft and used as a
sealing surface. One advantage of wear sleeves is that they can be easily re­
placed in case of surface distress. Seal manufacturers’ catalogs list standard
shaft diameters for which lip seals are readily available. The diameter tolerances
are as follows:
210 Gear Drive Systems

Shaft diameter (in.) Tolerance

Up to 4.0 ±0.003
4.001-6.0 ±0.004
6.001-10.0 ±0.005

The maximum eccentricity a lip seal can accommodate varies with the shaft rpm.
At 1000 rpm the eccentricity can be up to 0.015 in. while at 4000 rpm eccen­
tricity is limited to 0.007 in. Operating eccentricity is a combination of the
amount by which the shaft is off center with respect to the bore and the
dynamic runout. It can be measured by the total movement of an indicator
mounted on the casing and held against the side of the shaft while the shaft is
slowly rotated.

Bore Specification

The seal outside diameter-bore interface is a potential oil leakage path. The press
fit of the seal in the housing should be a minimum of 0.004 in. with higher press
fits for bore diameters above 4 in. These recommendations apply to ferrous housings.
If another material such as aluminum is used, the higher coefficient of thermal
expansion must be considered. The bore finish should be a maximum of 125 /iin.
It is possible to procure lip seals with elastomeric coatings on the outside
diameter which will seal at the bore interface, but these seals must be assembled
with higher press fits than seals with metal casings. It is good practice to brush a
synthetic rubber coating on the bore inside diameter prior to assembly to fill
minor imperfections on the bore and enhance sealing.

Lip Seal Materials

The most commonly used lip seal materials in gearbox applications are the
nitriles, sometimes known as Buna N. The standard nitrile compound is com­
patible with most mineral oils and can operate continuously at temperatures
from -65 to 225°F. Nitrile seals are relatively inexpensive and readily available.
'This material is not suited for temperatures above 250°F because it tends
to harden and it is not recommended for highly compounded lubricants
(EP additives).
For high-speed and high-temperature applications fluoroelastomers are
recommended. These materials, sometimes known by the trade name Viton, can
be used in a temperature range o f -4 0 to 400°F. They have outstanding resist­
ance to a wide variety of fluids, including synthetic lubricants such as
Mil-L-7808 and Mil-L-23699. For critical applications the fluoroelastomers are
Bearings and Seals 211

undoubtedly the best technical choice. The disadvantages of this material are
higher cost and limited availability of particular configurations.
Polyacrylate elastomers are used for operation with EP-type fluids. They
have good resistance to temperatures up to 300°F but have poor low-
temperature properties.
Silicones are applicable to a wide temperature range of -100 to 300°F and
also have low friction and wear properties. They have poor resistance to oxidized
oil and some EP additives.

Carbon Face Seals

Figure 4.43 illustrates a carbon face seal installation. The seal assembly consists
essentially of two flat-surfaced rings, the runner rotating with the shaft and the
stationary carbon member carried in the housing. The carbon sealing face must
be free to follow the motions of the runner despite axial shaft movement or
deviations from pure rotational movement. In order to accommodate shaft axial
and radial movement, yet maintain a sealing force, the carbon sealing face is
spring loaded against the runner. The spring force is basically constant over some
range of axial movement and the seal must be axially positioned at assembly to
be within its operating range. This may be accomplished by shimming at
assembly. Because the carbon.sealing face is free floating a secondary static seal
is required to close the leakage path and this is usually an elastomeric O-ring.
To achieve long seal life the spring force must be minimized yet adequate to
overcome axial friction, inertia, and dynamic forces. It can be seen from Figure
4.43 that at the radial sealing interface fluid tends to be pumped back into the
bearing cavity due to centrifugal force; thus the face seal can seal against small
pressure differentials.

Figure 4.43 Carbon face seal.

212 Gear Drive Systems

The rubbing surfaces must be compatible such that during mechanical

contact there is no galling, abrasion, or wear. Carbon is used because of its
rubbing friction characteristics, compatibility with many materials, temperature
resistance, low mass, low porosity, and ease of fabrication. Runners can be
hardened stainless steel or plated steels such as 4140 coated with chrome or
carbide. Runner hardness should be Rc 58 minimum and the runners should be
ground and lapped to a flatness of three helium light bands.
Normally in a face seal installation the runner is the rotating member;
however, there are some designs with stationary runners and rotating seal
assemblies. This type of design is speed limited because of the effect of
centrifugal force on the internal seal components.

REFERENCES

1. SKF Engineering Data, SKF Industries Inc., King of Prussia, Pa., 1973.
2. The Rolling Elements Committee of the Lubrication Division of the
ASME, Life Adjustment Factors for Ball and Roller Bearings, 1971
(Library of Congress Card Number 70-179492).
3. Grubin, A., and Vinogradova, I., Investigation of the Contact of Machine
Components, Moscow, Ts-NIITMASH, Book 30, 1949 (DSIR, London,
Translation 337).
4. Archard, J. and Cowking, E., Elastohydrodynamic Lubrication of Point
Contacts, Proceedings o f the Institution o f Mechanical Engineers, Vol.
180, Part 3b, 1965-1966, pp. 47-56.
5. SKF Catalog 310-110,,Spherical Roller Bearings, SKF Industries, Inc.,
King of Prussia, Pa., June 1981.
6. Palmgren, A., Ball and Roller Bearing Engineering, SKF Industries Inc.,
King of Prussia, Pa., 1945.
7. NACA Technical Note 3003, September 1953.
8. Benes, Capacity/Cost Factor Can Help You Find Bargains in Bearings,
Machine Design, November 2, 1972.
9. Conway-Jones, J. M., Application and Design of Plane Bearings in Power
Transmission Machinery, First International Power Transmission Con­
ference, June 1969.
10. Yahraus, W. A., Influence of Lubrication System Variables on Sleeve
Bearing Performance, SAE Paper SP-390, May 1975.
11. Egli, A. The Leakage of Steam Through Labyrinth Seals, ASME Paper
FSP-57-5, 1935.
12. Horve, L. A., Fluid Film Sealing, Elastomeric Lip Seals, ASLE Education
Course, Chicago Rawhide Manufacturing Co., Chicago, 1961.
5
LUBRICATION SYSTEMS

The purpose of a gearbox lubrication system is to provide an oil film at the con­
tacting surfaces of working components to reduce friction. Equally as impor­
tant, the lubrication system absorbs heat generated in the gearbox and dissipates
it so that component temperatures do not become excessive. Prior to discussing
specific details of gearbox lubrication, it is appropriate to review lubrication
fundamentals.

VISCOSITY

Friction between rubbing surfaces is reduced by separating the surfaces with a

film of oil. The lower fluid friction is substituted for the frictional resistance of
dry metal surfaces. Viscosity is a measure of fluid friction. Consider Figure 5.1,
where plate 1 is moving with a velocity U over a stationary plate 2 and the plates
are separated by an oil film of thickness h. The force F on plate 1 causes a
shearing stress in the fluid. The fluid in contact with the moving plate is at
velocity U and the fluid in contact with the stationary plate is at rest. Newton’s
law of viscous flow states that the shearing stress in the fluid is proportional to
the rate of change of the velocity, and if we assume that the rate of shear is
constant,

Ss = ^
s h
If U is in inches per second, h is in inches, and Ss is in pounds per square inch,
Mis the absolute viscosity in pound seconds per inch squared. A pound second

213
214 Gear Drive Systems

u PLATE 1
F

PLATE 2

Figure 5.1 Fluid friction.

per inch squared is called a Reyn in honor of Osborne Reynolds. In the metric
system the absolute viscosity is usually expressed in centipoise:
Z = p(6.9 X 106)
where
Z = centipoise or 1/100 dyne-sec/cm2
= Reyns or lb-sec/in.2
jjl

Quite often, viscosity is expressed in Saybolt seconds universal (SSU), where the
viscosity is determined with an instrument that measures the time in seconds for
60 cm3 of a fluid to pass through a standard capillary tube at a given tempera­
ture. This is a kinematic viscosity and is related to absolute viscosity as follows:

where
Zk = kinematic viscosity, cSt
T = number of Saybolt seconds
The absolute viscosity in centipoise is

where p is the specific gravity at the given temperature. Specific gravity is

measured by the petroleum industry at 60° F with a glass hydrometer calibrated
in degrees as specified by the American Petroleum Institute. This is an arbitrary
scale which is converted to actual specific gravity as follows:
141.5
^ 60 131.5 + degrees API at 60°F
Lubrication Systems 215

Table 5.1 Viscosity and Specific Gravity of SAE Oils

Say bolt seconds Specific
Degrees P
universal gravity, p
SAE API at at
number 100°F 210°F 60°F 60°F 100°F 210°F
10 183 46 30.2 0.875 0.861 0.822
20 348 57 29.4 0.879 0.865 0.827
30 489 65 28.7 0.883 0.869 0.830
40 680 75 28.3 0.885 0.871 0.832
50 986 90 26.6 0.895 0.881 0.842
60 1394 110 26.3 0.897 0.883 0.844
70 1846 130 25.6 0.901 0.887 0.848

where p60 is the specific gravity at 60°F. The specific gravity at any other tem­
perature is related to p60 as follows:
p = p 60 - 0.00035(°F -60)
Table 5.1 gives SSU and p60 values for various SAE oils. It can be seen from this
table that the viscosity of oil varies widely with the operating temperature. The
viscosity-temperature relationship of lubricating oils can be plotted as a straight
line if the scales are correctly arranged. To determine the viscosity of an oil at

Figure 5.2 Oil viscosity versus temperature. (From Ref. 1.)

216 Gear Drive Systems

any temperature the Saybolt seconds are found at two temperatures, usually 100
and 210°F and a straight line is drawn between these points. Figure 5.2 presents
the viscosities of some common fluids as a function of operating temperature.

VISCOSITY INDEX

The viscosity index is a measure of how much the oil viscosity varies with tem­
perature. It would be ideal if the fluid viscosity were the same at low and high
temperatures, but this is unattainable. Fluids that have low viscosity variations
with temperature have a high viscosity index, whereas a low viscosity index
defines a fluid with a widely fluctuating viscosity with respect to temperature.
Typical viscosity indexes for petroleum oils range from 90 to 110.

POUR POINT

The pour point of an oil is the lowest temperature at which it will flow. This
parameter must be considered when designing a unit for outdoor use; however,
in most low-temperature applications the oil will be heated prior to startup.
When no provision for oil heating is made, the pour point should be 20°F above
the lowest expected ambient temperature.

GEAR LUBRICANTS

The selection of a gear oil depends on several factors, including the unit’s
operating speed and load, ambient temperatures, and which lubricants will be
available at the operating site. The most important parameter in selecting a lubri­
cant is the viscosity. High-speed units require less viscous oil than gears operating
at low speed. At high speed an acceptable oil film is generated at the tooth con­
tact even with a low-viscosity oil. Also, the churning that occurs at high speed
will be less severe with a low-viscosity oil, resulting in lower power losses. At
lower operating speeds a thinner oil film is generated and more viscous oils are
required to separate the contacting surfaces. Also, low-speed gears are loaded
to higher levels. Often, a gear unit will contain both high- and low-speed gear
meshes. In this case a compromise must be struck and some development may
be necessary. AGMA Standard 250.04 [2] defines a series of oils by viscosity as
shown in Table 5.2. Each grade is given an American Gear Manufacturers Associ­
ation (AGMA) lubricant number.
Table 5.2 Viscosity Ranges for AGMA Lubricants

Viscosities of
Rust and oxidation Extreme pressure former
inhibited gear oils Viscosity rangea gear lubricants0 AGMA system^
mm2/s Equivalent
AGMA lubricant number (cSt) at 40°F ISO gradeb AGMA lubricant number SSU at 100°F
1 41.4-50.6 46 193-235
2 61.2-74.8 68 2 EP 284-347
3 90-110 100 3 EP 417-510

4 135-165 150 4 EP 626-765

5 198-242 220 5 EP 918-1122
6 288-352 320 6 EP 1335-1632

7 Compe 414-506 460 7 EP 1919-2346

8 Compe 612-748 680 8 EP 2837-3467
8A Compe 900-1100 1000 8A EP 4171-5098
Note: Viscosity ranges for AGMA lubricant numbers will henceforth be identical to those of ASTM 2422.
aViscosity System for Industrial Fluid Lubricants, ASTM 2422. Also British Standards Institute, B.S. 4231.
bIndustrial Liquid Lubricants—ISO Viscosity Classification. International Standard, ISO 3448.
cExtreme pressure lubricants should be used only when recommended by the gear drive manufacturer.
dAGMA 250.03, May 1972 and AGMA 251.02, November 1974.
eOils marked Comp are compounded with 3 to 10% fatty or synthetic fatty oils.
Source: Ref. 2.
Table 5.3 Specification for R&O Gear Oils (Including Compounded Gear Lubricants)
Property Test procedure Criteria for acceptance
Viscosity ASTM D88 Must be as specified
Viscosity index ASTM D2270 90 min
Oxidation stability ASTM D943 Time to reach a neutralization number of 2.0:a
AGMA grade Hours (minimum)
1,2 1500
3,4 750
5, 6 500
Rust protection ASTM D665 No rust after 24 hr with synthetic seawater
Corrosion protection ASTM D130 No. 1 strip after 3 hr at 120°C (250°F)
Foam suppression ASTM D892 Must be within these limits:
Max volume of foam (ml) after:
Temperature 5-min blow 10-min rest
Sequence I 24°C (75°F) 75 10
Sequence II 93.5°C (200°F) 75 10
Sequence III 24°C (75°F) 75 10
Demulsibility ASTM D2711 Must be within these limits :a
Max. percent water in the oil after 5-hr test 0.5%
Max. cuff after centrifuging 2.0 ml
Min. total free water collected during entire test 30.0 ml
Cleanliness None Must be free from grit and abrasives
^ h e criteria for acceptance indicated for oxidation stability and demulsibility are not applicable to compounded gear oils.
Source: Ref. 2.
Lubrication Systems 219

RUST- AND OXIDATION-INHIBITED OILS

AGMA lubricant numbers 1 to 8 are petroleum-based rust- and oxidation-

inhibited oils which meet certain American Society of Testing and Materials
(ASTM) specifications, as shown in Table 5.3:
Oxidation Stability (ASTMD943) In this laboratory test, pure oxygen
is bubbled through a mixture of oil and water in the presence of copper and iron
wire catalysts at 95°C. The test life is reported as the time in hours it takes the
oil to reach an acidity number of 2.0.
Rust Protection (ASTM D665) This procedure measures the rust-
preventing characteristics of an oil in the presence of synthetic seawater. A steel
specimen is used.
Corrosion Protection (ASTMD130) This test evaluates the ability of an
oil to control the corrosion of copper and copper alloys in the presence of water
at elevated temperatures.
Foam Suppression (ASTMD892) When air contaminates an oil circula­
tion system, the efficiency of the system depends on the oil’s natural resistance
to foaming and its ability to break foam quickly. In this test foam is created in
the test oil by blowing air through it for 5 min. The volume of foam is recorded
at the end of a 10-min settling period.
Demulsibility (ASTMD2711) The degree of oil demulsibility is deter­
mined by the amount of time it takes equal portions of distilled water and oil to
separate after the two have been mixed by a 1500-rpm steel paddle at a tempera­
ture of 180°F.
The lighter rust- and oxidation-inhibited oils, AGMA lubricant numbers 1
to 4, are sometimes referred to as turbine oils and are widely used in gear units,
particularly those operating at high speeds.

EXTREME PRESSURE GEAR LUBRICANTS

Extreme pressure lubricants are petroleum-based oils containing special chemical

additives which can increase the load-carrying capacity of gears by forming a
film on the metal surfaces which provides separation when the lubrication film
becomes thin enough for the asperities to contact. Some boundary films will
melt at lower temperatures than others and will then fail to provide protection
at the surfaces. For this reason, many extreme pressure lubricants contain more
than one chemical for protection over a wide temperature range. Some of the EP
additives commonly used in gear oils are those containing one or more com­
pounds of chlorine, phosphorus, sulfur, or lead soaps. EP additives are
Table 5.4 Specification for Extreme Pressure Gear Lubricants
Property
220
Test procedure Criteria for acceptance
Viscosity ASTM D88 Must be as specified
Viscosity index ASTM D2270 90 min
Oxidation stability ASTM D2893 Increase in kinematic viscosity of oil sample at 95°C (210°F) should not
exceed 10%
Rust protection ASTM D665 No rust after 24 hr with distilled water
Corrosion protection ASTM D130 No. 1 strip after 3 hr at 100°C (212°F)
Foam suppression ASTM D892 Must be within these limits:
Max volume of foam (ml) after:
Temperature 5-min blow 10-min rest
Sequence I 24°C (75°F) 75 10
Sequence I I 93.5°C (200°F) 75 10
Sequence I I I 24°C (75°F) 75 10
Demulsibility ASTM D2711 Must be within these limits: AGMA grades
(Modified for 90 ml 2EP to 6 EP 7EP, 8EP
water)
Max. percent water in the oil 1.0% 1.0%
after 5-hr test
Max. cuff after centrifuging 2.0 ml 4.0 ml
Min. total free water collected during 60.0 ml 50.0 ml
entire test (start with 90 ml of water)
Cleanliness None Must be free from grit and abrasives
EP property ASTM D2782 An oil must pass both a 60-lb Timken OK load, and 11 stages on the
(Timken Test) FZG machine with A/8.3/90°C parameters for acceptance.
DIN 51-354
(FZG Test)
Gear Drive Systems

Additive solubility None Must be filterable to 25 jdm (wet or dry) without loss of EP additive
Source: Ref. 2.
Lubrication Systems 221

chemically reactive and care must be taken when they are used that metals such
as zinc or copper which may be in the gear unit are not attacked. An existing
unit should not be changed from a straight mineral oil to an EP oil without the
manufacturer’s approval.
Lubricant numbers 2EP to 8EP in AGMA Standard 250.04 [2] are EP oils
which meet certain ASTM specifications, as shown on Table 5.4. Most of the
tests are as those described for Table 5.3, however, for EP properties a Timken
test and a FZG test are specified.
In the Timken test a rectangular steel block is brought into contact with
a rotating steel cylinder to determine the maximum load a lubricant will carry
before its film strength is exceeded. The load is increased until the block shows
surface distress, which indicates lubricant failure. The FZG test uses operating
gears to measure wear and surface distress. This is a back-to-back test with two
connected gears operating at 1760 rpm and two connected 2640-rpm pinions.
Load is applied by loosening the coupling in the pinion shaft and placing weights
on a load arm. The preload is locked into the system by tightening the bolts of
the coupling, and the load arm and weights are removed before starting the
machine. Twelve uniform weights are consecutively applied and after each run
the gears are checked for weight loss and visual condition. When a 10-mg weight
loss is recorded between two runs the oil is considered to have failed. The test
gears are spur and are immersed in 194°F oil during operation.

SYNTHETIC LUBRICANTS

Synthetic lubricants are a broad range of compounds derived from chemical syn­
thesis rather than from the refining of petroleum. They have the following
advantages:
High-temperature thermal and oxidative stability
Low-viscosity variation over a broad temperature range
Low-temperature capability
Long service life
Synthetic lubricants used in helicopter transmissions and geared gas
turbine engines have the ability to operate at temperatures as low as -65°F and
as high as 400°F or more during a duty cycle. These lubricants have been
developed for military applications and are often designated by military
specifications.
The largest class of synthetic lubricants in use today are the esters, which
are materials that contain the ester chemical linkage. Two oils that are com­
monly used in gear applications are Mil-L-23699 and Mil-L-7808. Esters are
characterized by an even balance of properties. They have wide operating
222 Gear Drive Systems

Table 5.5 AGMA Lubricant Number Recommendations for Enclosed Helical,

Herringbone, Straight Bevel, Spiral Bevel, and Spur Gear Drives
Ambient temperature It“P
-10 to+10°C 10 to 50°C
Type of unita and low-speed center distance (15 to 50°F) (50 to 125°F)
Parallel shaft (single reduction)
Up to 200 mm (to 8 in.) 2-3 3-4
Over 200 mm to 500 mm (8 to 20 in.) 2-3 4-5
Over 500 mm (over 20 in.) 3-4 4-5
Parallel shaft (double reduction)
Up to 200 mm (to 8 in.) 2-3 3-4
Over 200 mm (over 8 in.) 3-4 4-5
Parallel shaft (triple reduction)
Up to 200 mm (to 8 in.) 2-3 3-4
Over 200 mm, to 500 mm (8 to 20 in.) 3-4 4-5
Over 500 mm (over 20 in.) 4-5 5-6
Planetary gear units (housing diameter)
Up to 400 mm (to 16 in.) O.D. 2-3 3-4
Over 400 mm (over 16 in.) O.D. 3-4 4-5
Straight or spiral bevel gear units
Cone distance to 300 mm (to 12 in.) 2-3 4-5
Cone distance over 300 mm (over 12 in.) 3-4 5-6
Gearmotors and shaft-mounted units 2-3 4-5
f
High-speed units 1 2
aDrives incorporating overrunning clutches as backstopping devices should be referred to the
gear drive manufacturer as certain types of lubricants may adversely affect clutch
performance.
^Ranges are provided to allow for variation in operating conditions such as surface finish,
temperature rise, loading, speed, etc.
CAGMA viscosity number recommendations listed above refer to R & O gear oils shown in
Table 5.3. EP gear lubricants in the corresponding viscosity grades may be substituted where
deemed necessary by the gear drive manufacturer.
dFor ambient temperatures outside the ranges shown, consult the gear manufacturer. Some
synthetic oils have been used successfully for high- or low-temperature applications.
ePour point of lubricant selected should be at least 5°C (9°F) lower than the expected
minimum ambient starting temperature. If the ambient starting temperature approaches
the lubricant pour point, oil sump heaters may be required to facilitate starting and ensure
proper lubrication.
^High-speed units are those operating at speeds above 3600 rpm or pitch line velocities
above 25 m/sec (5000 fpm) or both. Refer to AGMA Standard 421, Practice for High Speed
Helical and Herringbone Gear Units, for detailed lubrication recommendations.
Source: Ref. 2.
Lubrication Systems 223

temperature ranges and high viscosity indexes (125 to 250). Thus esters require
low torques for low-temperature operation and provide good lubrication charac­
teristics at high temperatures.
A limitation of ester lubricants is low compatibility with some polymeric
materials such as those used in seals. Also, synthetic lubricants are significantly
more expensive than conventional petroleum oils.

LUBRICANT VISCOSITY SELECTION

In general, the lowest viscosity oil sufficient to form an adequate oil film at
all operating conditions should be chosen. As a guide, Table 5.5 gives recom­
mended lubricant numbers for enclosed drives. The viscosity associated with
each number is presented in Table 5.2. In practice, the lubricant selection is
usually a compromise between the requirements of the various oil-wetted
components such as gears and bearings and the particular application require­
ments. For instance, a turbine generator set packager may desire a common
lubrication system for the turbine, gearbox, and generator. If the turbine
requires a particular lubricant such as a synthetic oil, the gear manufacturer
will be requested to design for this fluid.

TYPES OF LU B R IC A TIO N SYSTEMS

There are two types of lubrication systems in use: splash and forced feed. In the
splash system the unit is filled with oil to a predetermined level and operated as
a sealed system with no external connections. In the forced feed system, oil is
introduced into the unit through jets under pressure. Scavenge oil is pumped
through a cooler and filter prior to reentering the gearbox.
The splash system is far simpler and less expensive than the forced feed
design but is applicable only to low-speed units. As speeds increase, the heat
generated in the gearbox becomes excessive and an external system is required
to cool the lubricant. Also, oil must be precisely introduced at the gear and
bearing interfaces, and this is accomplished through strategically placed jets.
For every gear drive there is a mechanical rating: the load the transmission
can transmit based on stress considerations. In addition, there is a thermal rating,
which is the average power that can be transmitted continuously without over­
heating the unit and without using special cooling. AGMA thermal ratings [3]
are based on a maximum oil sump temperature of 200°F. If the thermal rating
is less than the mechanical rating, additional cooling or a forced-feed lubrication
system is required.
224 Gear Drive Systems

An empirical method for estimating the thermal rating of low speed (maxi­
mum rpm of 3600 or pitch line velocity of 5000 fpm) gear units is given in
Ref. 3. Basically, the calculation of the thermal rating is a heat transfer problem
where assumptions must be made for the heat generated in the unit and the
coefficient of heat transfer of the surface area of the casing. If sufficient heat
can be conducted and radiated through the casing the gearbox sump temperature
will stabilize below the limiting temperature. Quite often gear units are located
in areas where conditions are detrimental to good heat transfer. The following
environmental conditions must be considered when defining the thermal
rating [3]:
Operation in an enclosed space
Operation in a very dusty atmosphere where fine material covers the gear unit
Operation in a high-temperature ambient such as near motors, turbines, or hot
processing equipment
Operation in high altitudes
Operation in the presence of solar or radiant energy
Auxiliary cooling can be used in combination with splash lubrication to
increase the thermal rating. Air is forced past the radiating surfaces of the gear
casing by strategically placed fans. The casing can also be cooled by a water
jacket. In this scheme water passages are built into the gear housing, usually on
the high-speed side, and heat is carried away by a cooling water flow. To operate
a gear unit at maximum efficiency, auxiliary cooling schemes should include
thermostatic controls so that the oil is not cooled unnecessarily. Operating with
too cool a lubricant increases churning losses. The heat transfer from the gear
casing can be increased by adding cooling fins which will increase the surface area.

Splash Lubrication

Successful splash lubrication of a gearbox is more a matter of development than

analysis. The problem is to distribute oil to each component sufficient for lubri­
cation and cooling yet minimize churning losses and heat generation. Figure 5.3
shows a two-stage parallel shaft unit using splash lubrication. The simplest
scheme is to let the bull gear dip into the oil, as shown in the upper part of
Figure 5.3, and lubricate all bearings and the high-speed mesh by the mist that is
created by the action of the bull gear churning through the oil. This scheme will
be more successful wtih antifriction bearings than with journal bearings, which
require far more oil. To provide a positive oil supply to the bearings a feed
trough can be incorporated, as shown in Figure 5.3, which catches oil flung off
the bull gear. The oil drips down the casing walls, is trapped by the trough, and
is distributed to the bearings.
A further degree of sophistication is shown at the bottom of Figure 5.3.
An oil baffle surrounds the lower portion of the bull gear, allowing the oil level
Lubrication Systems 225

to be raised such that the large gear on the intermediate shaft dips into the
lubricant and positively lubricates the high-speed mesh. Because the bull gear
cannot churn through the sump, heat generation is minimized and efficiency
increased. If the bull gear speed is so slow that sufficient oil is not splashed on
the casing walls to lubricate the bearings, scrapers can be used to strip the oil
ring that rotates with the bull gear rim. The oil is stripped from the rim and
directed to the bearing feed trough (Figure 5.4). A splash lubrication system
requires an oil drain, oil filler, an oil-level monitor and a breather. In cold
ambients an immersion heater should be provided in the sump. An opening
to the sump sufficiently large to allow cleaning of sludge after draining is
also good practice.
226 Gear Drive Systems

Forced-Feed Lubrication
Figure 5.5 illustrates a typical forced-feed lubrication system for a gearbox. The
shaft-driven oil pump (A) sucks oil from the tank (D) through the suction pipe
(I). From the pressure side of the oil pump the oil is directed through a cooler
(C) and filter (B). A pressure,relief valve (Q) is located at the inlet to the
gearbox to hold the feed pressure at a constant predetermined level. An auxiliary
pump (N) is incorporated to prime the system prior to starting. It could also be
used as a backup in case of failure of the main pump. Check valves (G) are
located such that the main pump does not pump through the auxiliary system
and the auxiliary pump does not pump into the pressure side of the main oil
pump. A bypass is provided at the cooler (H) which can be thermostatically
controlled so that the oil is not cooled to too low a level. At various points in
the system temperature and pressure sensors are located.
When designing a lubrication system the first step is to estimate the oil
flow to the components and the gearbox efficiency. The temperature rise across
the gearbox can then be calculated:

AT =
MCp
where
AT = temperature rise, °F
M = oil flow, lb/min {Note: 1 gpm = 7.5 lb/min)
Q = heat generated, Btu/min [Note: Q = hp(42.4)]
Cp = specific heat of oil, = 0.5 Btu/lb-°F
For example, a gearbox transmitting 1000 hp with 98% efficiency will reject 20
hp or 848 Btu/min of heat to the oil. If the gearbox flow is 20 GPM or 150
lb/min the temperature rise across the gearbox will be 11°F. In other words, if
the oil enters the unit at 130°F it will gain 11°F and be at a 141°F temperature
level in the sump. Some working areas of the gearbox may be at significantly
higher temperatures than the bulk sump temperature.
Lubrication Systems 227
P 0 0

D N

Key: A, oil pump; B, oil filter; C, oil cooler; D, oil tank; E, pressure gauge;
F, thermometer; G, check valve; H, overflow valve; I, suction pipe; K, pressure
pipe; L, return flow pipe; M, oil level gauge; N, primer and spare pump unit;
O, temperature and flow gauge; P, connection for remote thermometer;
Q, pressure regulator.
Figure 5.5 Forced-feed lubrication system. (Courtesy o f American Lohmann
Corporation, Hillside, N. J.)
228 Gear Drive Systems

A typical maximum oil in temperature limit for a gearbox with forced-

feed lubrication is 130°F. The cooler must be sized to cool the oil to this
temperature under the worst operating conditions, such as a hot day. A typical
oil temperature rise might be 30°F across the unit. These values are for mineral
oils; synthetic oils can operate at higher temperature levels. Some turbine gear
sets operate with oil in temperatures of 170°F.
The amount of oil flow supplied to a gear mesh is usually based on experi­
ence and experimental data. A rule of thumb might be 0.017 lb/min per horse­
power or 0.002 gpm/hp. This is a generous flow and should result in a low-
temperature rise. To optimize the efficiency of a gear unit a program can be
conducted, gradually reducing oil flow until a predetermined maximum tempera­
ture rise is reached. Reducing oil flow will always increase efficiency, par­
ticularly in high-speed gearing, where churning is significant, but at the expense
of higher scavenge temperatures.
The quantity of oil flow passing through the gearbox is regulated by the
oil jet diameters, journal bearing leakages, and oil feed pressure. Usually, gearbox
feed pressures are on the order of 20 to 100 psig. To hold a constant feed
pressure a regulating valve is incorporated at the gearbox inlet asshown in
Figure 5.5 A typical pressure regulating valve will have aspring-loadedbypass
loop which opens up when the design feed pressure is exceeded and directs
excess oil back to the sump.
The flow of oil through a jet can be calculated as follows:

Q = KA y 2(^ AP (W)(60)

where
Q = flow, lb/min
K = oil jet discharge coefficient
G = gravity acceleration, 386 in./sec2
AP = pressure drop across the jet, lb/in.2
W = oil specific weight, lb/in.3
A = jet area, in.2
If the discharge coefficient is assumed to be 0.65 and W is taken as 0.032 lb/in.3
the equation is
Q = 194A V a P = 152(D)2 v'AP
where D is the oil jet diameter inches. Note that oil flow varies with the square
of the jet diameter and as the square root of the feed pressure. The minimum
practical jet nozzle diameter is approximately 0.030 in. in diameter. Smaller jets
tend to clog too easily due to foreign materials in the oil stream. It is good prac­
tice to have at least two jets at each lubricating position in case one clogs up.
Lubrication Systems 229

Gears with long face widths must have oil evenly applied along the face
width. If oil is introduced unevenly, thermal distortions will cause uneven load
distribution. A long-face-width gear should have several jets spraying along the
axis. There are also special spray nozzles that can be used to fan the oil flow out
over a large area.
Relatively little oil is required for lubrication provided that it is properly
applied. The bulk of the oil flow is required for cooling the gear tooth and
blank. Thus it would appear logical to spray a small amount of oil at the
incoming side of a gear mesh and a large amount of oil at the outgoing side,
where it can do the most efficient cooling job. In practice, for low- and
moderate-speed gear applications lubrication on either the incoming or outgoing
side can be satisfactorily developed. For very high speed gears, above 20,000
fpm, it has been shown that lubrication on the outgoing side is more efficient
than on the incoming. This is because the bulk of the oil is not being dragged
through the mesh, experiencing churning losses. High-speed gear lubrication,
however, is very much an art and a particular application may require extensive
development to minimize thermal distortions and heat generation.
The extent to which the gear mesh is penetrated by the oil jet is a func­
tion of the velocity of the jet. The velocity in turn is a function of the feed
pressure. Industrial applications generally operate with feed pressures in the
order of 30 psig, whereas high-speed aerospace gearsets introduce oil at pressures
of 100 psig. There are no conclusive studies determining optimum feed pres­
sures; however, for high-speed gears it would seem that higher pressures would
be beneficial.
Figure 5.5 illustrates the components incorporated in a typical forced-feed
lubrication system. The following paragraphs describe in detail the major compo­
nents of such a system.

PUMPS

The gearbox oil pump must deliver a given quantity of oil over a wide range of
oil in temperatures and viscosities. At startup the oil will be cold and viscous.
During operation on hot days the lubricant will warm up to the maximum design
allowable value. Also, the pump must be capable of priming itself and over­
coming the pressure drops in the line between the reservoir and the suction
(inlet) port of the pump.
For gearbox lubrication rotary pumps are generally used. Figure 5.6 illus­
trates some widely used configurations. Rotary pumps are positive-displacement
devices which deliver a given quantity of fluid with each revolution. At any
particular speed the oil flow is practically constant regardless of the downstream
conditions. The pressure developed is dependent on the resistance of the
230 Gear Drive Systems

INLET OUTLET

Figure 5.6 Rotary pump configurations.

Lubrication Systems 231

discharge piping. The resistance includes the effect of oil jet orifices, pipe
friction, elbows, and so on. Allowable pump discharge pressure is limited by the
strength of the pump and the power of the driving unit.
Other types of pumps, such as centrifugal or piston, are not used on
gearbox applications, for the following reasons:
Centrifugal pumps develop pressure as a result of centrifugal force and are
used mostly where large flow volumes at relatively low pressures are
required. The flow delivered varies considerably with changes in dis­
charge pressure, and centrifugal pumps are not capable of self-priming.
Piston pumps are insensitive to discharge pressure and are self-priming; however,
due to their reciprocating motion and the inertia effects of the moving
parts, speeds are relatively low; therefore, pump capacities are low
compared to rotary configurations. Also, the flow tends to pulsate in a
piston pump configuration.
Rotary pumps in gearboxes can be flange mounted to the unit and driven
by one of the gearbox shafts or independently mounted with an electric motor
or other prime mover driving. When shaft driven, the flow will vary directly with
shaft speed. Commercial pump speeds rarely exceed 1800 rpm in gear, vane, or
lobe pump configurations, although some small aircraft types achieve much
higher speeds. There are screw pump designs which can achieve speeds to
approximately 5000 rpm. This is because the flow through the pump is axial and
fluid speed is relatively low compared to the peripheral speeds of gear or vane
pumps. Figure 5.7 is a schematic of a screw-type pump.

OIL INLET

OIL OUTLET

Figure 5.7 Screw pump configuration. (Courtesy of Delaval IMP Pump

Division, Trenton, N. J.)
232 Gear Drive Systems

The pump discharge pressure will depend on the required gearbox feed
pressure and the pressure drop between the pump outlet and gearbox inlet.
This pressure drop is a result of the losses in the filter, cooler, valves, lines, and
so on, in the system. The pressure drop can be considerable and care must be
taken that the pump design pressure is not exceeded. The great majority of
commercial pumps are designed for pressures of 150 psi maximum, although
much higher pressure capacity is possible. A pressure relief valve can be incor­
porated downstream of the pump outlet to protect the pump against over­
pressure in case of downstream blockage.
The pumping action of all rotary pumps is essentially the same. On the
inlet side a void is created as the pumping elements rotate. Fluid, forced by
atmospheric pressure, fills this space and is transported to the outlet side. In a
gear pump, as the gears rotate, fluid is trapped between the gear teeth and the
case, and is carried around to the discharge. In a vane pump, the rotating
member with its sliding vanes is set off-center in the casing. The entering fluid is
trapped between the vanes and the inside of the case and is carried to the outlet.
The term “positive displacement pump” means that with each revolution a
specific volume of fluid, depending on the geometry of the elements, is passed
through the pump. If there were no clearance between the rotating elements, or
between the rotating elements and the casing, the volume of fluid pumped could
be easily calculated and predicted. Clearance does exist, however, and depending
on the discharge pressure, there is always some internal leakage from the outlet
to the inlet side of the pump. This leakage or volumetric pump efficiency must
be considered when designing a pump for a specific application. Another con­
sideration is the range of lubricant viscosity the pump will experience. Maximum
internal leakage will occur when the fluid is at minimum viscosity; thus the
capacity of the pump must be sized for this condition. A gearbox oil pump
should be specified to be oversized at least 15% as far as flow requirement is
concerned to account for operating variables and pump deterioration over time.
To control the flow into the gearbox a pressure regulating valve is usually
incorporated at the unit inlet. The valve is set to maintain a constant inlet
pressure, which if the gearbox orifices are correctly sized, will result in the
design flow. As the oversized oil pump delivers more flow than required, the
pressure will exceed the set point and the regulating valve will bypass enough
flow to maintain the design pressure into the gearbox. Figure 5.8 shows the flow
versus speed characteristic of a shaft driven oil pump with a pressure relief valve.
From startup to point A the flow varies directly with speed. At point A the
pressure, which varies with the square of the speed, has reached the set point of
the relief valve and the valve cracks open. The flow into the gearbox is then regu­
lated at a constant amount with the valve bypassing sufficient flow to keep the
feed pressure constant.
Lubrication Systems 233

Figure 5.8 Pump flow versus speed characteristic using pressure relief valve.

A significant parameter in designing the lubrication system is the pump

suction requirement. Fluid flows into a pump due to the difference in pressure
between the pump inlet and fluid source. In most cases the fluid source is at
atmospheric pressure, so there is 14.7 psia pressure available to force the lubri­
cant into the pump. Tending to retard the lubricant flow are the static lift
(maximum level of fluid below the pump inlet), friction losses due to pipes,
valves, elbows, and so on, and entrance losses from the reservoir to the pipe and
the pipe into the pump inlet. Commercial rotary pumps generate suction lifts of
approximately 5 to 15 in. Hg; 30 in. Hg is equivalent to 14.7 psia; thus com­
mercial pumps can withstand an inlet line pressure drop of approximately 2.45
to 7.35 psia. If the pressure drop in the inlet line is excessive, the pump will not
fill with oil and as a result oil flow and discharge pressure will be erratic and
drop off. Also, air bubbles will be formed in the fluid and as they implode,
cavitation damage may occur, with the potential of pump failure.
To minimize inlet line pressure drop, the reservoir should be located as
close to the pump as possible and not too far beneath it. Line and fitting sizes
should be as large as possible and bends, valves, fittings, and so on, should be
minimized. To help pump priming at startup, the inlet port to the pump should
be full of oil. If it is possible for oil to drain down from the inlet port during
shutdowns, a foot valve can be placed in the inlet line at a low point. This type
of valve has a light spring load which keeps it closed at shutdown and blocks
234 Gear Drive Systems

oil drainage from the pump inlet. At startup the pump suction creates a pressure
differential across the valve, opening it and allowing flow. In any system when
starting up after a long shutdown period it is a good idea to check the pump
inlet port to ensure that it is full of oil.
A frequent problem in pump inlet systems is air leaks. Small leaks that
are not easily detectable allow air into the line. The air displaces lubricant and
the pump discharge pressure and flow will drop off and become erratic. It is
difficult to find these leaks but they are quite common and must be considered
when troubleshooting pump problems.
The power required to drive an oil pump can be estimated as follows:

P = ------ 9 .AP q j l in-3/gal)--------- = 0 0006QA

(12 in./ft) (33,000 ft-lb/min/hp)
where
P = horsepower
Q = flow, gpm
Ap = pressure developed, psi
This is the theoretical power, to which must be added mechanical and viscous
losses. The mechanical losses include friction drag of all moving parts, such as
bearings, seals, and so on. Viscous losses include the power lost to fluid viscous
drag against the pump components as well as shearing of the fluid itself. The
mechanical efficiency of a pump is a measure of the magnitude of the horse­
power to loss to the theoretical horsepower.
Gearboxes in critical applications often have a shaft-driven main oil pump
and an electric motor-driven auxiliary pump. The auxiliary pump is used for pre­
lubrication prior to start up and as a backup in case of failure. Suitable valving must
be designed in the system (Figure 5.5) so that the pump flows are properly directed.

FILTRATION

Gearbox lubrication systems are subject to contamination due to a variety of

causes. The internal components wear generating particles washed away by the
oil stream. Also, at assembly, during maintenance, and even during operation
foreign particles find their way into the unit, and these contaminants, if uncon­
trolled, will cause wear and possibly failure of bearings or gears. Filtration is the
mechanism that captures particles in a fluid by passing them through a porous
medium. The degree of filtration is generally specified by a micrometer rating:
jum = 10"4 cm = 0.00003937 in.
As a point o f reference a human hair is 70 /im or 0.0028 in. thick.
Lubrication Systems 235

Typical gearbox filtration specifications call for 40 /im or 0.0016 in.

filtration, which can be interpreted to mean that the filter will allow no particle
greater than 40 (jtm to pass. When one considers that journal bearing film thick­
nesses are on the order of 0.001 in. and less, and gear mesh film thicknesses
lower, a case can be made for specifying finer filtration than 40 jum, possibly 10
or even 3 jum. It has been proven that fine filtration is beneficial to component
life; however, to obtain full benefits great care must be taken with the entire
system or the filter will clog very quickly or be in bypass. For instance, a 55-gal
drum of new oil can contain a billion particles 10 jum or larger. Thus new oil
should be filtered prior to use in the gear unit. Also, great care must be taken in
cleaning all gearbox components and piping prior to assembly. The unit should
be flushed with an auxiliary filter prior to use with the standard filter.
The absolute filtration rating of a filter is the diameter of the largest hard
spherical particle that will pass through under specified test conditions. To test
a filter, a measured quantity of contaminant, typically spherical glass beads in
the range 2 to 80 jum diameter, are passed through the test article and captured.
The captured fluid is then passed through a very fine membrane filter which is
examined under a high-powered microscope. By visual observation, the diameter
of the largest glass bead on the membrane is determined.
Filters are sometimes rated by efficiency. One such rating is the nominal
efficiency, which is a measure of the retention by weight of a specified artificial
contaminant that is passed through a test filter and captured by a membrane
filter. Nominal efficiencies are typically in the range 90 to 98%; however, this
rating method is losing favor due to lack of reproducibility and uniformity.
Gaining favor in the industry is the beta (j3) filtration rating, which is the
rating of the number of particles greater than a given size in the influent fluid
to the number of particles greater than the same size in the effluent fluid passed
through a test filter. For instance, if the number of particles greater than 10 jdm
per unit volume of fluid entering a filter is 5000 and the number of particles
greater than 10 fim per unit volume of fluid leaving the filter is 50, the j310
rating is

U,„ = ^ = 100
50
The j3 rating can be converted to an efficiency as follows:
Px - 1
Ex = ------
Px
where
Ex = efficiency expressed as percentage of the filter medium’s ability to
remove particles over a particle size x, by count
Px = beta filtration ratio (fi rating) for contaminants greater than x jum
236 Gear Drive Systems

In the previous example,

E 10 - --------------
100 = 99%
meaning that the filter will remove a minimum of 99% of all particles greater
than 10 /xm. The following table gives efficiencies for various beta ratios:

P Ratios versus Efficiency

P Ratio Efficiency (%)
1 0
2 50
20 95
50 98
100 99
1,000 99.9
10,000 99.99

Figure 5.9 shows how the beta ratio concept can be used. A p3 filter with a
ratio of 1000 will remove 99.9% of all particles greater than 18.5 jum. A p10
filter with a ratio of 1000 will remove 99.9% of all particles greater than
26 jum. The P10 filter will remove 99.7% of all particles greater than 18.5 jum.
The beta ratio is determined using a continuous flow of contaminated
fluid through a test filter. Samples of the fluid from upstream and downstream
of the filter are taken and analyzed with an automatic particle counter. This
laboratory instrument is capable of determining the particle size and distribu­
tion per unit volume of fluid.
Filter media are usually fibrous materials comprised of many fine fibers
randomly oriented with diameters ranging from 0.5 to 30 /im. Materials most
commonly in use are cellulose, cotton, micro-fiberglass, and synthetics. The
smaller the diameter of the fiber used, the closer they can be compacted and
the higher the filter efficiency.
When designing a filter, in addition to the filtration rating, the following
points must be considered:
Flow rate
Fluid type
Operating temperature
Operating pressure
Dirt-holding capacity
Pressure drop
 Lubrication Systems 237
BETA RATIO

15 20
PARTICLE SIZE ( y m )

Figure 5.9 Typical beta ratios for j33 and j810 depth-type media versus particle
size. (From Ref. 4.)

The flow rate will determine the line size and clean pressure drop for a given size
of filter. Fluid type, operating temperature, and pressure will dictate the con­
struction and materials of the filter components. Dirt-holding capacity is an
indication of how long an interval may be expected between filter changes.
As the filter collects contaminants during operation, the pressure drop
across the filter will increase. Filter elements have collapse pressure ratings and
the pressure drop cannot be allowed to exceed this value. Other system con­
siderations such as exceeding the maximum oil pump discharge pressure will
also affect the filter pressure drop allowed. In order to monitor the pressure
238 Gear Drive Systems

differential across the filter, pressure gauges can be incorporated on the inlet and
outlet sides. Filters are available with integral differential pressure indicators that
can provide a warning when a set point is reached. To protect the filter and
provide a continuous flow of oil, some filters incorporate a bypass valve which
will open when a predetermined pressure drop is reached. At bypass, only a
portion of the fluid will pass through the filter, allowing unflltered lubricant to
enter the system. For critical applications dual oil filters with changeover valves
are specified. With such a design one filter can be serviced while the system is
operating using the other filter. Filter elements can either be cleanable and
reusable or disposable. Cleanable filter elements are usually made of wire mesh
and cleaning is commonly accomplished in an ultrasonic liquid bath. For
gearbox applications, in terms of filtration quality, reliability, and ease of
maintenance, it is recommended to use disposable elements.
The oil filter should be located on the pressure side of the pump down­
stream of any component that might produce contaminants, such as the oil
cooler. Filters placed in the pump suction line can cause inlet problems as the
pressure drop across the filter increases with use.

COOLERS

In a forced-feed lubrication system, the oil in temperature to the gearbox is

controlled by passing the hot scavenge oil through a heat exchanger. In order
to specify a cooler, the maximum expected oil scavenge temperature must be
estimated and the maximum allowed oil temperature into the unit defined.
The cooler must be capable of achieving the required oil temperature when
exposed to the maximum ambient air temperature anticipated in the applica­
tion. A generous safety margin should be applied during design to account for
deterioration of the cooler. For instance, over time, a water-oil cooler will
experience deposits in its tubing reducing efficiency.
The two types of coolers used are liquid to liquid or liquid to air.
Figure 5.10 schematically illustrates an oil-water cooler. The hot oil entering
through the shell side encounters the cooling water and an equilibrium of tem­
peratures occurs, cooling the oil and heating the water. Where water is not
available, radiators are used, blowing cooling air over oil tubes. Air-to-oil
coolers require larger envelopes than water-to-oil coolers. Also, on hot days,
the air temperature will limit the amount of cooling a radiator can achieve.

OIL RESERVOIR

The oil reservoir or tank may be integral with the gearbox (wet sump) or
separately mounted (dry sump) and connected to the gearbox by piping. If
Lubrication Systems 239

WATER SIS!

Figure 5.10 Typical oil-water cooler configuration.

integral, on high-speed gearing, the sump should be isolated from the gear
windage by some type of baffling so that the oil in the reservoir is still. The level
of oil in the reservoir will vary from a maximum when the unit is shutdown and
oil has drained from lines and components to the minimum allowed during
operation. The minimum operating level is determined by the length of time it
is desired to have the oil dwell in the tank to effect deaeration. For instance,
if the nominal flow is 20 gpm and a dwell time of 2 min is specified, the mini­
mum oil volume should be 40 gal. The longer the dwell time, the more time air
entrained in the oil has to settle out. It is difficult to determine how much
dwell time a particular application will require by analytical means. If there is
the potential for significant amounts of air to be entrained in the oil, significant
dwell time may be required. Air usually enters the lubrication system through
pressurized labyrinth seals. Reference 5 specifies an 8-min retention (dwell)
time based on nominal flow and total volume below the minimum operating
line. This is a conservative value. If there is no likelihood of significant aeration,
2 min of dwell time or less may be sufficient.
240 Gear Drive Systems

After setting the minimum operating level of the oil reservoir an estimate
is made of the anticipated oil loss due to leakage and any other reasons and the
time interval between oil additions in order to determine the maximum
operating level to which the reservoir will be filled. At shutdown, when lines and
components such as coolers and filters drain back into the reservoir, the oil level
will be higher than the maximum operating level; therefore, the tank must have
sufficient volume to accommodate the drain backflow and still retain some air
space at the top. At initial startup some quantity of oil is required to fill all lines
and components. It is good practice to run for a brief period and then check the
oil level to determine if additional lubricant is required to come up to the proper
operating level. Oil levels are best monitored by a sight glass and dipsticks are
also used.
To ensure complete drainage for cleaning and oil changes the bottom of
the reservoir should slope toward a low-point drainage connection. The oil pump
suction line should connect slightly above the high end of the reservoir bottom
so that sediment on the bottom is not pulled into the pump inlet line. Oil return
lines should be piped into the reservoir above the maximum operating level away
from the area around the pump suction connection so that the oil around this
area is undisturbed. To facilitate inspection and cleaning of the reservoir, suffi­
ciently large openings must be provided.
When cold ambient operation is anticipated, the reservoir can be heated
either by applying external heat or incorporating a thermostatically controlled
immersion heater in the tank. With immersion heaters care must be taken not to
overheat the oil in contact with the element, since this will lead to lubricant
degradation. The heater watt density should not exceed 15 W/in.2 [5].

BREATHER (VENT)

The gearbox breather is used to vent pressure that may be built up in the unit.
Pressure may occur as a result of air entering the lubrication system through
seals or the natural heating and cooling of the unit. When a cold gearbox starts
up, the heat generated during operation causes air within the case to expand.
The same effect is noted with units operating outdoors as they heat up during
the day and cool at night.
Breather caps usually have some type of baffling built in to keep par­
ticulate contamination out of the gearbox, but moisture can enter. It is possible
to have a completely sealed system and incorporate an expansion chamber in
place of the breather. Figure 5.11 presents a schematic of an expansion chamber
which is screwed into the gear case or reservoir. The expansion or contraction
of air is accommodated by flexing of the diaphragm. In harsh environments
care must be taken in choosing the diaphragm material such that it is compatible
with the environment.
Lubrication Systems 241

In a complicated system the gearbox vent may be inadvertently routed to

an area with an unfavorable environment and cause operating difficulties. For
instance, if the gearbox is vented to a pressurized area, the gearbox may be
back-pressured, resulting in oil leakage through the seals. If vented to an area
under a vacuum, oil might be sucked out of the unit through the vent.

PIPING

Although apparently simple, the piping connections for a gearbox can be a

source of aggravation at assembly and startup. In many cases the responsibility
for supplying piping and lubrication system components is split between the gear
manufacturer and the user. The purchase order should be specific as to who
supplies what, which specifications apply, and where the interfaces are in order
to avoid complications at installation.
As far as gearbox oil feeds are concerned, it is good practice to have only
one external connection with all other oil passages placed inside the casing. Then
if there is any slight leakage in the piping connections, it will be internal and
harmless. Also, there will be less chance of damage to the piping during shipping
and assembly. When units are horizontally split, the lubrication piping and
components should be mounted on the bottom of the casing such that the top
half can be disassembled without disconnecting any lubrication lines.
The piping arrangement must be carefully designed to minimize pressure
drops. Care must be taken to position components such as coolers and filters
such that they will not drain at shutdown and require refilling at each startup.
There are several methods of fabricating lubrication system piping. Reference 5
specifies bending the pipes to suit the required contours and welding them at
connections. This is the best way to ensure that there will be no leakage; how­
ever, it is more expensive and requires more skill than using threaded joints. The
piping terminates in flanged connections which are bolted to the mating part.
Seamless carbon steel piping to ASTM A106 or ASTM 192 is recommended.
242 Gear Drive Systems

Figure 5.12 Flared fitting.

Figure 5.13 Flareless tube fitting.

Lubrication Systems 243

There are two basic types of threaded joints. One is the pipe thread, which
is tapered and produces a metal-to-metal seal by wedging surfaces together as the
pipe is screwed in. When connecting this type of joint, sealant should be applied
to the male thread. Avoid the first two male threads from the end to keep sealer
out of the system. Tetrafluoroethylene-fluorocarbon tape is not recommended
to seal pipe threads since pieces may break off and enter the lubrication system,
clogging the oil jets. SAE straight thread fittings depend on an O-ring for sealing.
Care must be taken in assembly or the O-ring will be damaged, resulting in oil
leakage. More positive in stopping leakage than threaded fittings are tube
fittings, which fall into three groups: flared, flareless, and the previously men­
tioned welded or brazed. A typical flared fitting consists of three pieces (Figure
5.12). The seal occurs when the soft tubing is pressed against the hardened
conical seat of the connector body. Standard flare angles are 37° and 45° and
should never be mixed.
Flareless fittings are usually used where tube thicknesses make use of
flared tubes difficult. Figure 5.13 illustrates a bite-style flareless tube fitting
where the wedging action of the ferrule provides the seal.

REFERENCES

1. E. R. Booser and R. C. Elwell, Keeping Lubricants Flowing at Low Tem­

peratures, Machine Design, March 24, 1977, pp. 74-79.
2. AGMA Standard 250.04, Lubrication of Industrial Enclosed Gear Drives,
American Gear Manufacturers Association, Arlington, Va., 1981.
3. AGMA Standard 420.04, Practice for Enclosed Speed Reducers or
Increasers Using Spur, Helical, Herringbone and Spiral Bevel Gears,
American Gear Manufacturers Association, Arlington, Va., December 1975.
4. Filtration Manual, Purolator Technologies, Newbury Park, Calif., 1979.
5. API Standard 614, Lubrication, Shaft Sealing and Control Oil Systems for
Special Purpose Applications, American Petroleum Institute, Washington,
D.C., September 1973.
6
MATERIALS AND HEAT TREATMENTS

The selection of the gear material and its heat treatment is the most basic deci­
sion in the design process. The strength of a gear tooth, as shown in Chapter 3 is
proportional to its hardness; therefore, the size, configuration, and cost of a gear
unit are dependent on the choice of gear material and its processing. Although
there are a wide variety of materials and processes to choose from, practical con­
siderations tend to narrow the options. By this is meant that at any gear manu­
facturer’s facility the type of equipment available for tooth finishing and heat
treating will limit the designer’s choices. In fact, the equipment available dictates
the design philosophy and the material and processing specifications.
The types of gearboxes discussed in this book usually use gears made of
alloy steels. These materials carry the greatest load in terms of power trans­
mitted per pound of gear and offer high reliability. Steel is defined as iron with
carbon percentages in the range of 0.15 to 1.5. A steel is considered to be an
alloy when the maximum of the range given for the content of alloying elements
exceeds one or more of these limits; 1.65% Mn, 0.60% Si, or 0.60% Cu; or when
a definite range or minimum amount of any of the following elements is
specified: aluminum, chromium, cobalt, columbium, molybdenum, nickel,
titanium, tungsten, vanadium, or zirconium. Table 6.1 lists the chemical com­
positions of some of the most widely used gear steels. The American Iron and
Steel Institute (AISI) or Society of Automotive Engineers (SAE) numbers
designate the steel composition and alloy type. The last two digits signify the
carbon content: for example, AISI 4140 contains 0.40% carbon. The first two
digits designate the approximate alloy content.
Gears are generally either through-hardened in the range Rc 32 to 43
(Bhn 300 to 400) or surface hardened in the range Rc 55 to 70. The hardness

245
246 Gear Drive Systems

Table 6.1 Chemical Compositions of Gear Steels

AISI number C Mn Ni Cr Mo
4140 0.38/0.43 0.75/1.0 0.80/1.0 0.15/0.25
4340 0.38/0.43 0.60/0.80 1.65/2 0.7/0.9 0.20/0.30
4620 0.17/0.22 0.45/0.65 1.65/2 0.20/0.30
4320 0.17/0.22 0.45/0.65 1.65/2 0.4/0.6 0.20/0.30
8620 0.18/0.23 0.70/0.90 0.40/0.70 0.4/0.6 0.15/0.25
3310 0.08/0.13 0.45/0.60 3.25/3.75 1.4/1.75
9310 0.08/0.13 0.45/0.65 3/3.5 1/1.4 0.08/0.15
2317 0.15/0.20 0.40/0.60 3.25/3.75

range between Rc 43 and Rc 55 is seldom used since the steel is too hard to cut
in this state. Therefore, it may as well be fully hardened to obtain maximum
load capacity and finish ground if necessary.
The strongest and most durable gear meshes are made up of two surface-
hardened gears using the carburizing process. Utilization of surface-hardened
gears will result in the minimum gearbox size, creating savings in materials,
machining, and handling costs. On the other hand, there are increased costs asso­
ciated with carburized gears. Because of heat-treat distortion, griding after
hardening will be necessary to achieve high precision of the gear teeth. The
grinding and heat treating equipment is very costly and may not be readily
available. From a technical point of view, however, carburized, hardened, and
ground gears are the best alternative for the high-speed, high-power drives. Of
the materials listed in Table 6.1, the low-carbon steels—4620, 4320, 8620, 3310,
9310, and 2310—are carburizing grades.
Another surface-hardening process that is used extensively is nitriding.
This process produces a very hard, wear-resistant case which is somewhat more
brittle than that of a carburized gear; thus the fatigue properties are not as good.
Heat-treat distortion resulting from nitriding is not as great as that from car­
burizing; thus it is sometimes possible to use gears after nitriding without a
grinding operation. Even if grinding is planned, nitriding is sometimes used to
limit distortion. It is not uncommon to mesh a carburized and ground pinion
with a nitrided gear. When the gear is large, grinding is often considered
impractical either because equipment is not available or simply because the cost
is too high. AISI 4340 and 4140 are nitriding steels.
Another common material combination is a surface-hardened pinion
meshing with a through-hardened gear. Although the gear strength and durability
is considerably lower than that of the pinion, the gear will experience fewer
stress cycles and a reasonable fatigue life can be achieved. The hardened pinion
may tend to wear the softer gear in improving mesh characteristics and also cold
Materials and Heat Treatments 247

working the gear, which increases its surface hardness. The through-hardened
gear is considerably cheaper than a surface-hardened gear since the heat treat­
ment is much simpler and no grinding is required.
The least expensive mesh combination is a through-hardened pinion
mating with a through-hardened gear. The size of such a unit will be large com­
pared to a gearbox incorporating surface-hardened gears: as much as twice the
envelope and weight; however, where there are no restrictions on size or inertia,
this type of design is widely used. Most double helical units have through-
hardened gears. Of the materials listed in Table 6.1, AISI 4140 and 4340 are
through-hardening steels.
There are other methods of surface hardening in addition to those men­
tioned above, such as induction or flame hardening. Also, other materials are in
use, such as AISI 1040,1045, 1137, and 1340. These are plain carbon steels
where the alloying element content is low. These materials and processes are
economical and satisfactory for gears requiring only a moderate degree of
strength and impact resistance. Where cost is the primary consideration, these
options should be considered.

HARDENING FU N D A M E N TA LS

Steel, as received from the mill, has a structure of nearly pure iron (ferrite) and
iron carbide (cementite). To harden a part it has to be heated in the range of
1450 to 1600°F, where it takes the form of austenite, a structure of iron and
carbon stable at high temperatures. Depending on how quickly the part is
cooled, the austenite transforms to other structures. Figure 6.1 illustrates the
formation of the various structures of carbon steel due to quenching. If cooled
very rapidly, a hard, strong, and somewhat brittle form called martensite, which
is carbon dissolved in iron, is formed.
The maximum hardness that any steel can attain after quenching is a
function of its carbon content. Figure 6.2 illustrates this point. The curve
reflects 100% transformation of austenite to martensite. From the curve it can
be seen that the through-hardening, high-carbon steels 4140 and 4340 can attain
a maximum hardness of Rc 58 after quenching. Gears are not used in this state
since a fully quenched martensitic structure lacks the touchness to be able to
resist shock loading. To improve the gear tooth’s ductility and toughness its
structure is modified by tempering, a controlled reheating of the part to a tem­
perature below the austenitizing temperature. This process also reduces residual
stresses left in the gear during hardening. Tempering temperatures as low as
300°F can produce significant increases in toughness and ductility and reduc­
tions in residual stresses with small hardness decreases.
248 Gear Drive Systems
FERRITE CEMENTITE

HEAT

AUSTENIT E

INTERMEDIATE VERY S L O W
RAPID CO O L CO OL S L O W COOL

MARTENSITE FINE PEARLITE PEARLITE SPHEROSDITE

& MARTENSITE

Figure 6.1 Formation of steel structures during quenching.

CARBO N(% )

Figure 6.2 Relationship of carbon content to hardness.

Materials and Heat Treatments 249

Figure 6.3 Reduction of steel hardness due to tempering.

In the case of carburizing steels which initially have low carbon levels, the
tooth surface is enriched with carbon to approximately 0.85% and thus can
attain a surface hardness of approximately Rc 65 maximum. The core will
remain relatively soft, about Rc 35; thus the carburized tooth has a strong hard
case with a tough ductile core. This structure has been found to be optimum for
load transmisision. The carburized gear, like the through-hardened gear, must be
tempered after hardening. Figure 6.3 shows the surface hardness drop-off of an
AISI 9310 carburized gear at various tempering temperatures. The surface
hardness was Rc 65 in the as-quenched condition.

HARDENABIL1TY AND THE USE OF ALLOY STEELS

An alloy steel is chosen on the basis of the mechanical properties required in the
part and the heat treatment available to develop these properties. The charac­
teristic of a steel that will determine if it is acceptable is called hardenability.
Hardenability is a measure of the severity of cooling conditions necessary to
achieve the hardness required. Hardenability of an alloy steel is established by
the carbon and alloy content. To obtain a given hardness, higher-hardenability
steels require less severe quenching than do low-hardenability steels. A less severe
quench will result in lower distortion of the part and less chance of residual
stresses, which can manifest themselves in cracks during processing or gear
operation.
250 Gear Drive Systems

For the same section thickness a steel with good hardenability can be
hardened throughout with a far less severe quench than a low-hardenability steel.
For instance, a 1-in.-diameter AISI 1045 steel bar fully hardened at the outside
when quenched in water will have a hardness drop-off at its center of 5 points
Rc. A 1 in. diameter AISI 4340 bar will have negligible hardness drop-off
through its cross section after water quenching.
In order to achieve lower cooling rates and deeper hardening properties,
alloys are incorporated in the steel formulation. The following elements are
added to alter the properties of steel:
Carbon. The principle hardening element in steel
Manganese. A strengthening and hardening element
Nickel. Improves low-temperature toughness and ductility, reduces distortion
and cracking during quenching, and improves corrosion resistance
Chromium. Increases hardenability and forms hard, wear-resistant carbides
Molybdenum. Improves hardenability

W ATER Q UENC H

DISTANCE FROM Q UENCHED E N D - 1 / 1 6 IN.

Figure 6.4 Jominy test for hardenability.

Materials and Heat Treatments 251

The standard method of determining a steel’s hardenability is the Jominy bar

test. A vertically suspended bar 1 in. in diameter and 4 in. long is heated above
its critical temperature into the austenitizing range. It is then water quenched at
one end and air cooled at the other end. Hardness is measured along the bar and
the gradient between the fully hardened water-quenched end and the softer air-
cooled end represents the hardenability of the steel. If the steel has high harden­
ability, the hardness will be at a higher level farther from the water-quenched
end than that of a low-hardenability steel. Figure 6.4 illustrates the results of a
Jominy bar test.

STEEL QUALITY

The AISI designation of a steel specifies only the chemical composition of the
material. The quality in terms of freedom from seams, cracks, folds, inclusions,
or nonhomogenous structure may not be adequately controlled when specifying
a material in this manner. When the best quality is required, steels should be pro­
duced by the electric furnace process rather than the basic open hearth or
oxygen process. Macroscopic, microscopic, and magnetic particle inspection
should be required to determine material structure and cleanleness. For critical,
high-reliability applications it may be necessary to specify that material be pro­
duced by vacuum or double vacuum melting to assure adequate material
cleanliness.
The Society of Automotive Engineers issues a series of Aerospace Material
Specifications (AMS) that control the quality of steels. For instance, AMS 6260

Table 6.2 AMS Specifications for Gear Steels

AISI (SAE) AMS
designation designation Remarks
4140 6382H
4340 6414B Premium quality, cons. elec. vacuum melted
4340 6414H
4620 6294D
4320 6299B
8620 6276D Premium quality, cons. elec. vacuum melted
8620 6277B Premium quality, cons. elec. melted
3310 6250G
9310 6260H
9310 6265D Premium quality, cons. elec. vacuum melted
9310 6267B Premium quality
2317 -
252 Gear Drive Systems

covers air-melt AISI 9310 steel and AMS 6265 is the vacuum-melt version. The
AMS numbers corresponding to the steels in Table 6.1 are given in Table 6.2.
The steel melting process is critical in controlling material cleanliness. Air
melt is the standard procedure. Premium-quality steels undergo vacuum
degassing where the billet is heated to a temperature slightly below the melting
point in a vacuum environment and hydrogen and oxygen are removed from the
steel. This process minimizes nonmetallic inclusions such as aluminum oxide and
silicon oxide in the material. A step further in producing clean steel is to melt
the material one or more times in a vacuum so that impurities are removed as
gases. It has been shown that vacuum melt steels yield significant improvement
in gear reliability by minimizing the possibility of fatigue crack initiation due to
material impurities. Specifying an AMS steel may add to the cost of a gear
initially but in many instances can save the manufacturer the cost of scrapping
a semifinished part when material defects are uncovered during machining
operations.

PROCESSING OF THROUGH-HARDENED GEARS

Steel, as obtained from the mill, either in an as-forged or rolled condition may
not be uniform with high- and low-hardness areas. The first metallurgical opera­
tion to be performed is annealing, which involves slow heating, austenization of
the steel, and then slow furnace cooling. Annealing will improve machininability
by providing a uniform low-hardness structure. This facilitates the initial rough
machining.
The part may then be normalized, which involves heating above the
austenitizing temperature and then air cooling at an intermediate rate. The
resulting structure is somewhat harder, stronger, and less ductile than when in
the annealed state. The gear is finish machined at this point except for the teeth
and bearing journals. The normalized structure facilitates precise machining
and results in a good surface finish.
The part is then hardened and tempered to the design requirement. At this
point the gear teeth are cut and the bearing journals finished. Another temper is
performed to relieve machining stresses and possibly a finishing operation is
required on the teeth such as shaving or lapping. The journals might also require
re finishing.

CARBURIZING

In the carburizing process carbon is diffused into the surface of the teeth by con­
trolled exposure at temperatures of 1650°F or above for the length of time
Materials and Heat Treatments 253

necessary to achieve the desired case depth. After hardening the tooth has a high
surface hardness, decreasing to a specified core hardness which depends on the
carbon and alloy content of the particular steel used. Carburizing produces the
strongest tooth with respect to bending and pitting resistance by generating
residual compressive stresses in the case area. The hard surface has excellent wear
and scoring resistance.
Before carburizing, the gear teeth are cut and other areas that are to be
hardened are machined. Areas of the part that will remain soft must be insulated
from the carburizing medium. This may be accomplished by copper plating or
other masking procedures or by leaving excess stock to be machined after heat
treating. In some cases a carburized area can be machined after hardening if it is
to remain soft. After carburizing the part is hardened by quenching. The process
will cause distortion in the part; therefore, to attain precise tooth geometry the
gear teeth must be ground after hardening. It is important to minimize the
distortion so that the stock removal after hardening is small since the grinding
process is removing the hard case that is desired. In many cases the gear must be
quenched in a die so that distortion can be kept within acceptable limits.
During the carburizing process carbon is introduced in liquid or gaseous
form. The most common process now in use is gas carburizing, where a con­
trolled gas surrounds the part in a sealed furnace. The amount of time required
to carburize a part is dependent on the temperature; however, carburizing above
1800°F tends to coarsen the grain size to an unacceptable level.
The optimum carbon content to achieve maximum hardness is 0.080 to
0.090%. This is achieved by controlling the richness of the carbon medium. The
case depth is controlled by taking sample coupons out of the furnace at intervals
during the cycle. When the coupon achieves the required case depth the carburiz­
ing cycle is stopped.

Table 6.3 Recommended Case Depth versus Diametral Pitch for Carburized
Gears
Light case Standard case Heavy case
Diametral pitch (in.) (in.) (in.)
4 0.041 0.053 0.064
6 0.030 0.040 0.049
8 0.028 0.032 0.040
10 0.020 0.027 0.034
20 0.010 0.014 0.018
Source: Ref. 1.
254 Gear Drive Systems

Figure 6.5 Case depth measurement.

The effective case depth is defined as the depth at which the hardness of
the case is Rc 50. The case depth must be sufficiently deep to withstand the
stresses created by the transmitted load, yet not so deep that the tip of the tooth
becomes brittle with a tendency to break off. For this reason it is recommended
that tooth tips and end faces be masked off during carburizing and left at core
hardness. The case depth required will vary according to the tooth size or
diametral pitch. Table 6.3 gives recommended case depths for various diametral
pitches.
The case depth is measured by microhardness-testing a cross section of a
tooth. This requires sacrificing a production piece or including a test section of
gear teeth with the heat-treated lot. The case depth measurement is accom­
plished in the laboratory using an instrument which measures the hardness from
the surface inward every 0.001 in. The hardness survey should be made at the
tooth midheight perpendicular to the tooth surface. The root fillet region should
also be checked since it is harder for the carbon to penetrate this closed-in area.
Figure 6.5 illustrates the areas of the tooth where the survey should be taken.
It is not uncommon for the root hardness and case depth to be somewhat lower
than the flank, but both should be within design requirements.
In addition to the case depth, the engineering drawing should specify the
following:
Materials and Heat Treatments 255

Minimum case hardness or range. The maximum carburizing surface hardness

attainable after grinding is approximately Rc 63; therefore, the most
stringent practical drawing requirement is Rc 60 to 63. In many cases
Rc 58 minimum is specified and for moderately loaded gears Rc 55
minimum may be acceptable.
Core hardness range. Core hardness should be held within the range Rc 30 to
40. Higher core hardnesses may make the tooth too brittle, and lower
values have insufficient strength. It is difficult to control core hardness
closely since this parameter is dependent on the base steel carbon content,
which varies from one heat value to another. The process controls are
directed toward achieving case hardness and the core hardness falls where
it will.
There are variations possible in the process used to carburize gears which
involve a trade-off between cost and the quality of the metallurgical structure
obtained. Table 6.4 shows a carburizing process for a critical 30-tooth, 10-pitch
gear with a 1.0-in. face width. This is an elaborate process designed to achieve
the best possible metallurgical structure. Note that following carburizing there is
a subcritical annealing step the purpose of which is to refine the case structure.
This step is sometimes omitted and the part is cooled down to 1500°F and
quenched directly from the carburizing furnace. If this is done, care must be
taken during the cooling to surround the gear with a carbon atmosphere so that
decarburization of the part does not occur. A similar problem arises in the sub-
critical annealing step. If an air atmosphere is used, the part must be copper
plated so that scale formation does not occur during annealing or decarburiza­
tion during the subsequent hardening.
The hardening process involves heating the gear until the part is com­
pletely austenitized and quenching in an oil bath at 80 to 140°F. The objective
of the quench is to transform the austenite into martensite; however, in practice,
all the austenite can never be transformed. This is undesirable since the retained
austenite is not stable and if present at too high a level can spontaneously cause
distortion in the part. In order to transform as much of the retained austenite as
possible, carburized parts are sometimes subzero cooled in the range -100° to
-150°F. This cooling must be accomplished within 20 min of quenching, to
avoid austenite stabilization. The minimum volume of retained austenite that
can be practically achieved is 10%. For applications that are not extremely
critical, 20% can be allowed. Retained austenite is measured definitively by
x-ray diffraction or (visually estimated by) metallographic examination of an
etched cross section.
Carbides in the case structure are desirable for increased strength and
wear resistance, however, if they form continuous or semicontinuous networks
at the grain boundaries, stress concentrations result which can initiate crack.
256
Table 6.4 Carburizing Process for AISI 9310 Steel
Time at Quench
Process Temperature temperature Type temperature
operation (°F) (hr) atmosphere Type of Quench (°F) Remarks
Normalize 1725 2 ENDOa Chamber R.T.b
Harden 1500 1 ENDO Oil 130 5 min in oh
Temper 1050 4 Air Air R.T.
Carburize 1700 As required Carbon — - Furnace cool to 1500°F,
potential air cool to R.T.
Subcritical 1150 2 Air
anneal
Harden 1500 1 ENDO on 130 Agitated oil
Freeze -120 1 Liquid Thaw to R.T.
Temper 300 4 Air
aEndothermic atmosphere.
bR.T., room temperature.
Source: Ref. 1.
Materials and Heat Treatments 257

Figure 6.6 Typical metallographic standards for undissolved case carbides.

(From Ref. 1.)
255 Gear Drive Systems

Figure 6.6 illustrates the levels of severity of grain boundary carbides. These are
photographs of an etched cross section at 500X magnification. Figure 6.6A
shows scattered grain boundary carbides and is the maximum condition accept­
able for critical applications. Figure 6.6B illustrates a semicontinuous grain
boundary carbide network. This is a marginal condition and might be acceptable
for noncritical gearing. Figure 6.6C is a continuous grain boundary carbide
network and is not acceptable.

N1TR1DING

Gas nitriding is a surface-hardening process in which the part is heated to

approximately 1000°F in a furnace with an ammonia atmosphere. Quenching is
not required, the hardness developing as a result of the formation of hard
nitrides near the surface. AISI 4340 and 4140 are nitriding steels, their
chromium and molybdenum alloys combining with the nitrogen to form
nitrides. The steel must be hardened and tempered before nitriding to a temper­
ing temperature higher than the subsequent nitriding temperature. The nitriding
cycle is quite long. For instance, a 0.020-in. case depth requires approximately
40 hr of processing.
Because the part is not quenched and the process is performed at a rela­
tively low temperature, little distortion occurs during nitriding. For this reason
many gears are finished by cutting, shaving, or grinding in the soft state and then
hardened by nitriding. For extreme accuracy grinding after nitriding is still
required; however, for some large gears the process is still preferred to carburiz­
ing because of the limited distortion.
During the nitriding process an outer white layer is formed on the gear
teeth which is hard and brittle and therefore undesirable. With special process­
ing the white layer can be held to a maximum depth of 0.0005 in. which is
usually acceptable. Thicker white layers are unacceptable because of the possi­
bility of spalling or flaking. They must be removed either by grinding, some
other mechanical process or chemical action. Nitriding produces shallow case
depths compared to carburizing; therefore, if grinding after nitriding is required,
stock removal must be closely controlled to preclude grinding away the case.

FO R G ING OF GEAR BLANKS

Gears up to approximately 6 in. in diameter can be fabricated from bar stock.

Larger gears are forged from bars or billets. Forging is defined as the plastic
deformation of a metal at an elevated temperature into a predetermined size
or shape using compressive forces exerted through some type of die by a
Materials and Heat Treatments 259

AS ROLLED BAR S T O C K FOR GE D ST O CK

Figure 6.7 Grain flow due to forging.

hammer, press, or an upsetting machine. Many gear blanks are made of upset
forgings, where a round bar is struck axially between two flat dies. The bar is
shortened and the metal, following the path of least resistance, flows outward.
A forged part will have flow lines which are composed of fibrous non-
metallic inclusions or segregated phases which have been flowed in the direction
of working. These fibers, if properly oriented, add to the strength of a part.
Figure 6.7 shows the grain glow in a part made from bar stock compared to the
same part made by upset forging. The direction of the forged flow lines will
improve the bending resistance of the gear that will be fabricated from the bar.

HOUSING MATERIALS

In the design phase of a gearbox the question arises s to whether the housing
should be cast or fabricated. A fabrication is comprised of steel plates which are
welded or bolted together to form the casing. Usually, a carbon steel such as
AISI 1020 is used.
If only one or two units are to be manufactured, the economics favor
fabrication. In order to use a casting a pattern must be developed and the cost
of fabricating a pattern is significant. Also, if the program is under time pressure,
a fabrication can be made in less time than required to develop a pattern and
pour a casting.
As the quantity of gearboxes required increases, it becomes more eco­
nomical to go to a casting. The pattern costs are amortized over the production
quantity and the basic material cost of cast iron is lower than steel. Also, less
material is used since the casting can be shaped to conform closely to the gear
configuration.
260 Gear Drive Systems

There are technical advantages to a casting in that the gearbox weight can
be reduced and the casting provides more flexibility to the designer to achieve
noise and vibration attenuation.
Cast irons are alloys of iron and carbon containing more than 2% carbon.
The types of cast irons most commonly used for gear casings are gray iron and
ductile iron. They are specified by American Society of Testing and Materials
(ASTM) standards, which define chemical composition and physical properties.
Gray cast iron is graded by minimum tensile strength. For instance, a
class 30 gray iron would have a minimum tensile strength of 30,000 psi. Gray
iron grades range from 20 to 60. Ductile iron is specified by a three-part system.
For instance, an 80-55-06 alloy has a minimum tensile strength of 80,000 psi,
a minimum yield strength of 55,000 psi, and 6% elongation in 2 in. Ductile iron
classes range in minimum tensile strength from 60,000 to 120,000 psi.
The type of cast iron selected for a given application must take into
account not only the physical properties desired but the size and shape of the
casting. The solidification rate, which depends on casting geometry and foundry
practice, profoundly influences the resulting strength. Actual properties of a cast
part will vary with the cooling rate, cross section, microstructure, and the grade
of iron used. Like steel parts, cast irons are heat treated for a variety of reasons.
These reasons can be summarized under the headings of stress relief, annealing
and normalizing, and hardening and tempering.
The stress relief treatment is used because cast irons are susceptible to
growth when allowed to stand at room temperature for long periods of time.
Stress relief is accomplished at temperatures of 700 to 1300°F. If time permits,
castings can be aged for 6 months or more to allow growth to occur and the
stress relief can be omitted.
Annealing and normalizing heat treatments are used to modify the castings
metallurgical structure to increase machinability. Hardening and tempering, as
in steel parts, are applied to improve component strength.
In the casting process discrepancies sometimes occur which leave voids in
the material. These can sometimes be repaired by welding. Welding reworks
should be inspected by x-ray techniques to ensure a completely crack-free repair.

HARDNESS TESTING

Because so much attention is paid to gear tooth hardness it is appropriate to

review the definition of hardness and how this material property is measured.
Hardness of a material can be defined as its resistance to permanent deforma­
tion. There are several methods to measure this property, the most widely used
being Brinell and Rockwell.
Materials and Heat Treatments 261

In the Brinell test, a load, usually 3000 kg, is applied to the test piece by
means of a hardened 10 mm steel or tungsten carbide ball. The resulting impres­
sion is measured and the Brinell hardness number (kg/mm2) is calculated by
dividing the load applied by the area of the impression. The Rockwell hardness
test uses a diamond cone to indent hard materials. First a minor load of 10 kg
is applied to force the penetrator below the surface of the material, following
which a major load varying between 60 and 150 kg depending on the scale used
is applied. The hardness, which is proportional to the depth of penetration, is
read directly. The Rockwell machine is more flexible and easier to use than the
Brinell method; therefore, it is widely used for routine testing and the inspection
of heat-treated parts. There are two types of Rockwell machines, the Normal
tester for relatively thick sections, and the Superficial tester for materials too
thin to be tested with the normal instrument. Minor loads on the superficial
tester are 3 kg and the major loads vary from 15 to 45 kg. A cone-shaped
diamond indenter is used. For work with gear steels the A, C, and D scales on
the Normal tester are used. On the Superficial tester the 15-N and 30-N scales
are used.
In the laboratory microhardness testing is accomplished using two
different types of diamond indenters. One is a pyramid with a square base
[diamond pyramid hardness (DPH)] and the other is a rhombic-based pyramid

Table 6.5 Approximate Comparison of Hardness Scales

Rockwell
Brinell 10-mm
ball, 3000-kg 15-N C 30-N
load (15-kg load) (150-kg load) (30-kg load) DPH Knoop
614 90 60 77.5 695 732
587 89.25 58 75.5 655 690
560 88.5 56 74 617 650
534 87.5 54 72 580 612
509 86.5 52 70.5 545 576
484 85.5 50 68.5 513 542
460 84.5 48 66.5 485 510
437 83.5 46 65 458 480
415 82.5 44 63 435 452
393 81.5 42 61.5 413 426
372 80.5 40 59.5 393 402
352 79.5 38 57.5 373 380
332 78.5 36 56 353 360
313 77 34 54 334 342
297 76 32 52 317 326
283 75 30 50.5 301 311
262 Gear Drive Systems

(Knoop). Loads applied are on the order of 5 to 100 kg and the hardness
number is proportional to the ratio of the load to a characteristic length of the
impression to the second power. The test specimen is an etched, mounted cross
section of tooth and the dimensions of the impression left by the indenter are
read with a microscope. The DPH numbers correlate well with Knoop results up
to a hardness of approximately Rc 58. Beyond this level a Knoop reading of a
given specimen referred to the Rockwell C scale will yield a value 1 or 2 points
lower than a DPH reading of the same specimen. Table 6.5 gives approximate
hardness comparisons for Brinell, Rockwell, DPH, and Knoop. Hardness tests are
usually not performed on tooth flanks since the indentation can be a source of
stress concentration. On some designs there may be hardened areas which are
not functional and can be surfaces for hardness checking. In many cases the
gear tooth hardness is verified only by the microhardness survey of the cross­
sectioned tooth sample taken from the heat-treated lot. There is nothing wrong
with this practice and it is the best way to confirm hardness and case depth in
the tooth root.

N O N D E S TR U C TIV E TESTIN G

Nondestructive tests are used to detect mechanical defects in material and varia­
tions in condition or composition. These tests are performed on production
parts and do not impair the component’s function.

M A G N ETIC PARTICLE INSPECTION

This inspection method is used to identify discontinuities such as cracks, inclu­

sions, or pores which are at or near the surface of a ferromagnetic material. Such
discontinuities are areas of stress concentration and can propogate under cyclic
loading leading to component failure. Cracks can be a result of poor grinding
process or heat treating; therefore, it is good practice to provide for magnetic
particle inspection at various stages in the manufacturing cycle. When a defect
occurs it is best to identify it as early as possible so that minimum machining
time is spent on discrepant material.
Magnetic particle inspection is accomplished by magnetizing a part by
passing an electric current through it. Leakage magnetic fields occur at surface
or near surface discontinuities. These defects are revealed by their attraction
for finely divided magnetic particles which are introduced on the surface of the
part being inspected.
The widely used Magnaflux method applies iron oxide particles suspended
in kerosene or oil just before or while the part is being magnetized. This is called
Figure 6.8 Magnetic particle indications (indicated by arrows).
263
264 Gear Drive Systems

the wet method. There is a dry method where powdered iron oxide is applied
while the part is under magnetization. The iron oxide particles cluster about the
discontinuity and reveal its approximate location, size, and extent. The black
particles are sometimes colored red or gray for better contrast on dark objects.
There is a Magnaglo method of inspection where the particles are coated with a
fluorescent material which glows when activated by ultraviolet light. This
improves the contrast between the iron oxide particles and the part under
investigation.
It is important to demagnetize any part that has undergone magnetic
particle inspection. If gear teeth or shafts have any residual magnetism, they will
tend to attract machining chips or abrasive particles, which can cause difficulty
during operation.
To determine whether a part is acceptable following magnetic particle
inspection, acceptance and rejection standards must be defined. In general, no
indications of discontinuities are allowed on critical areas such as gear or spline
teeth. Also, indications that extend over or into an edge, chamfer, corner, radius,
fillet, or hole are not acceptable. Often, magnetic particle indications can be
removed by localized grinding or remachining. For instance, indications at the
edge of a tooth might be removed by regrinding a larger edge radius.
In noncritical areas some indications may be accepted. A typical specifica­
tion may be as follows:
A maximum of three indications 3/16 in. long or less are allowed per
noncritical area.
Figure 6.8 presents some examples of magnetic particle indications.

L IQ U ID PEN ETR A NT INSPECTION

Liquid penetrant inspection is a nondestructive method for finding discontinu­

ities that are open to the surface of solid and essentially nonpourous materials.
In gearboxes the process is usually used for the inspection of castings. The
principle of the liquid penetrant inspection method is to wet the surface of the
workpiece with a uniform liquid coating which migrates into cavities that are
open to the surface.
A widely used method is the Zyglo inspection, which uses nonmagnetic
particles that fluoresce under ultraviolet light. The particles are suspended in
oil and applied to the part to be inspected. Excess liquid is removed from the
part and the material remaining in any cracks or discontinuities is revealed under
inspection by ultraviolet light.
It is difficult to set standards of acceptance for discontinuities found in
castings since the shapes are usually unique to the particular application. The
location of the defect must be taken into account, and whether it may be
Materials and Heat Treatments 265

Figure 6.9 Surface temper indications.

266 Gear Drive Systems

detrimental to component strength or serviceability. Minor discontinuities in

noncritical areas can be accepted without rework. When there is some question
as to whether the discontinuity might propagate and cause distress, it can be
removed by localized machining.

SURFACE TEMPER INSPECTION

When hardened components are ground there is the possibility of overheating

the surface being machined and locally tempering the overheated areas. This may
occur due to improper grinding prac-tice or interruption of coolant flow during
machining. If the surface temperature in a localized area exceeds the tempering
temperature of the material, the area will soften. In the worst case the local
temperature reaches a level sufficiently high to reharden the material, resulting
in an area of brittle untempered martensite which can be a source of pitting
during operation.
To detect surface temper burns the component is dipped in an acid solu­
tion which etches the metal, exposing tempered areas. The process is usually
known as Nital etch. Unburned areas of an etched part will be uniformly gray
with a dull nonreflective surface. Tempering burns will appear as dark gray or
black areas on the etched part. The darkness of the color reflects the severity of
the burn. Rehardening burns will appear as white or light areas surrounded by a
black tempered area.
In order to set acceptance criteria for tempering, critical areas of the part
being inspected must be defined. Also, examples of light temper distinguished
by coloration must be available to use as comparisons. Critical gear areas are the
loaded flanks of the gear teeth and the root fillet area. Clutch surfaces and
bearing surfaces in rolling contact are also considered critical. No temper burns
of any degree are allowed in the critical areas. Light temper covering less than
20% of a noncritical area may be acceptable. Rehardening burns are not accept­
able in any area.
Because temper is a surface discrepancy, burns can sometimes be removed
by remachining the tempered area. In addition to exposing temper, the Nital
etch will show up areas that are deficient in carbon or have had excessive stock
removal during grinding. These areas will appear light in color when compared to
a normal etched surface. Figure 6.9 illustrates tempered gear teeth; the darker
color showing the tempered area, which in this case exhibited spalling.

REFERENCE

1. AGMA Standard 246.01 A, Recommended Procedure for Carburized Aero­

space Gearing, American Gear Manufacturers Association, Arlington, Va.,
November 1971.
7
MANUFACTURING METHODS

There are two basic methods of manufacturing gear teeth: the generating process
and the forming process. When a gear tooth is generated, the workpiece and the
cutting or grinding tool are in continuous mesh and the tooth form is generated
by the tool. In other words, the work and the tool are conjugate to each other.
Hobbing machines, shaper cutters, shaving machines, and many grinders use this
principle.
When a gear tooth is formed, the tool is in the shape of the space that is
being machined out. Some grinding machines use this principle with an indexing
mechanism which allows the gear teeth to be formed tooth by tooth. Broaches
are examples of form tools that machine all the gear teeth simultaneously. Gears
of the type discussed in this book are initially cut on a hobbing machine or
shaper cutter. The tooth forms are further refined by lapping, shaving, grinding,
or honing. Following are descriptions of these various processes.

HOBBING

Figure 7.1 illustrates a hobbing machine and Figure 7.2 presents the hob itself.
In this process the gear teeth are generated with the hob and workpiece rotating
in a constant relationship while the hob is being fed into the work. Hobbing is a
versatile and economical method of cutting gears. A hob of any given normal
pitch and pressure angle will cut the teeth of all spur and helical gears having the
same normal pitch and pressure angle. Hobs producing involute gears are basic­
ally straight sided and generate the involute form on the gear tooth by the
meshing action.

267
Figure 7.1 Hobbing machine. (Courtesy of Barber Colman Corporation,
Rockford, 111.)

Figure 7.2 Gear hobs. (Courtesy of Barber Colman Corporation, Rockford, 111.)
Manufacturing Methods 269

(AN

Figure 7.3 Shaper cutting machine. (Courtesy of Fellows Corporation, Spring­

field, Vt.)
270 Gear Drive Systems

Hobbing produces gears of qualities up to American Gear Manufacturers

Association (AGMA) Quality Class 12. To achieve high accuracy the hobbing
machine must be in excellent condition, the tooling holding the gear rigid and
accurate, the gear blank accurate, and the hob of high quality. All types of
materials can be hobbed. The great majority of hobbed gears are in a hardness
range Rc 30 to 38; however, with special material cutters hardnesses up to Rc 60
can be machined. Hobbing will produce surface finishes in the range 16 to
64 juin. rms.
The only limitations of the hobbing process are the inability to machine
internal gears and a requirement for axial space behind the gear teeth to allow
the hob to run out. Gears that are adjacent to a shoulder or cluster gears cannot
be hobbed.

Figure 7.4 Shaper cutters. (Courtesy of Fellows Corporation, Springfield, Vt.)

Manufacturing Methods 271

SHAPING

Figure 7.3 illustrates a shaper cutter machine and Figure 7.4 shaper cutters.
Shaping is a generating process where the tool is in the form of a shape conjugate
to the tooth being cut. When cutting involute gear teeth the shaper cutter is in
the form of an involute gear which is hardened and has cutting clearance on the
tooth sides. The gear blank and the cutter are rotated in the proper ratio while
the cutter reciprocates axially through the gear blank. If a spur gear is being
generated, the cutter will reciprocate through the workpiece in a straight path.
To generate a helical gear the cutter must reciprocate in a helical motion which
is imparted by a helical guide. This additional tooling required to cut helical
gears is a disadvantage of shaper cutting compared to hobbing.
Shaping can be applied to internal as well as external spur or helical gears.
Also, herringbone gears are cut by the shaping process. The shaper cutter does
not require a large runout beyond the gear; therefore, it is a good method for
cutting cluster gears or gears close to a shoulder. As with hobbing, the best
shaper cutting can produce gears up to AGMA Quality Class 12 with finishes to
16 juin. rms. Hardnesses to Rc 43 can be shaper cut; however, the great majority
of gears are cut in the range Rc 30 to 38.

LAPPING

Lapping is a method of correcting small errors in profile, lead, spacing, or runout

of gear teeth. A gear can be lapped either by running it with its mating gear or
meshing it with a lapping tool in the form of a gear. An abrasive lapping com­
pound is introduced into the mesh to promote removal of metal. The abrasive
medium must be uniformly spread across the teeth in an oil- or water-soluble
base and should contain a rust inhibitor. Gear laps are usually made of cast iron
into which the abrasive will embed.
For most lapped gears the process is short, typically minutes. Only small
beneficial changes can be attained by lapping and too lengthy a cycle will usually
destroy the profile. Because there is no sliding at the pitch line, lapping will tend
to remove material above and below the pitch line only. An auxiliary axial
sliding motion is incorporated in lapping machines to overcome this problem.
Wide face helical and double helical gears are often lapped for long periods
of time such as several hours to improve their tooth contact. It has been found
with wide face gears that have a helical overlap greater than 2 that no axial
reciprocating motion is necessary while lapping.
272 Gear Drive Systems

Figure 7.5 Shaving cutter. (Courtesy of National Broach and Machine Division
of Lear Siegler, Detroit.)
Manufacturing Methods 273

SHAVING

Shaving is a finishing operation which uses a high-speed-steel, hardened and

ground, precision cutter which is in the form of a helical gear (Figure 7.5). The
cutter teeth have gashes which act as cutting edges. Shaving will improve the
tooth spacing, profile, lead, runout, and surface finish which was generated in
the hobbing or shaping process. Shaving is often used to refine the tooth surface
prior to hardening, thereby minimizing heat-treat distortion.
The shaving cutter is meshed with the work gear in a crossed axis relation­
ship (Figure 7.6), and rotated while the center distance between the two is
reduced in small increments. Simultaneously, the work is traversing back and
forth in relation to the cutter.
Shaving can achieve gear qualities of up to AGMA Class 13. The quality of
a shaved gear is strongly dependent on the quality of the preceding hobbing or
shaping operation. Shaving can remove approximately 70% of the errors in the
as-cut gear. A surface finish of 25 juin. rms is normally achieved, with much finer
finishes possible.The optimum gear tooth hardness for the shaving process is Rc30.

Figure 7.6 Shaving cutter machine. (Courtesy of National Broach and Machine
Division of Lear Siegler, Detroit.)
274 Gear Drive Systems

Shaving may be applied to spur and helical gears, both external and
internal. Tooth profile modifications and crowning can be achieved. Profile
modifications are accomplished by grinding the correct form into the cutting
tool teeth. Crowning is accomplished by sinking the cutter more deeply into the
tooth ends than into the middle.

GRINDING

The most accurate gears are produced by the grinding process. It is used pri­
marily to finish hardened gears in the range Rc 55 and up. Grinding can produce
AGMA Qualities of Class 14 and higher and surface finishes as fine as 10 /bn.
rms. Gear tooth profile modification and crowning of the gear tooth face can be
accurately produced by the grinding process. The two basic techniques used to
grind gear teeth are the form grinding method and the generating method. Either
method, when closely controlled, can produce the highest-quality gears.
Figure 7.7 illustrates a form grinding machine which utilizes a disk-type
grinding wheel. The wheel form is contoured by a diamond dressing tool into the

Figure 7.7 Form grinding machine. (Courtesy of National Broach and Machine
Division of Lear Siegler, Detroit.)
Manufacturing Methods 275

form of the tooth space being machined out. Both the root and the tooth flank
area can be ground, or the flank area alone, by cropping the tip of the grinding
wheel. The wheel grinds one tooth at a time and an index plate having the same
number of spaces as the number of teeth in the gear is used to index around the
part. Tooth spacing is a function of the accuracy of the index plate and spacing
error can be minimized by using an index plate larger in diameter than the part
being ground.
Generating grinders use disk wheels of various types, such as conical or
saucer shaped or threaded worm types of wheels. The work is rolled with respect
to the grinding wheel and the wheel reciprocates axially with respect to the
work. Generating grinder wheels act as though they were straight-sided racks in
mesh with the gear being ground. The disk is therefore dressed in a straight­
sided form with the proper pressure angle. The straight-sided form is modified
to achieve whatever tooth modifications are required on the gear to be ground.
When grinding case-hardened gears it is important to control all phases of
the process in order to minimize the amount of stock removal during grinding.
Too much stock removal will result in loss of the hardened surface area and the
benefits of case hardening. This is particularly critical in the root fillet area.
In order to have sufficient case depth in this critical bending stress region, the
tooth must be cut prior to hardening with a generous root fillet radius. Figure
7.8A shows the shallow case depth that results in the root fillet area when the
radius is too small. The surface area available for carbon to penetrate in this
region is insufficient. After hardening, when the tooth is ground, little or no
case will be left and bending failure resistance will be significantly reduced. In
order to attain the largest possible root fillet radius the hob or shaper cutter
tips must be designed to produce an undercut below the form diameter of the
gear tooth (Figure 7.8B). This configuration is referred to as protuberance cut

Figure 7.8 Hardened gear teeth in the as-cut condition.

27(5 Gear Drive Systems
GROUND PROFILE
CUT PROFILE

CUSP

A B C

Figure 7.9 Hardened gear teeth after grinding.

since in order to generate the undercut the cutter has a protuberance added at its
tip. The case depth in the root fillet is greater with protuberance cut teeth
because there is a greater surface area available for carbon to penetrate during
the carburizing cycle. Another advantage of protuberance cutting is that only
the tooth flank need be ground, leaving the carburized case in the root fillet
intact. Figure 7.9 shows the types of tooth profiles that result from this grinding
technique. Assume that the as-cut undercut is 0.003 in., as shown in Figure
7.8B. If the stock removal during grinding of the tooth flank is 0.003 in., the
profile will be tangent to the root fillet radius (Figure 7.9A). This is the ideal
case. Stock removal of less than 0.003 in. will leave some undercut in the
finished tooth (Figure 7.9B) and stock removal greater than 0.003 in. will result
in a cusp (Figure 7.9C).
There is some question whether a tooth with unground root or one with a
fully ground root is stronger. The fully ground root may have a better form,
however, some of the hardened case has been removed. With a fully ground root
there is always the danger that stock removal has been excessive. Both grinding
methods are commonly used and, if correctly done, produce gears that achieve
successful operation.
The grinding operation is strongly influenced by the condition of the gear
after heat treatment. In carburized parts excessive case carbon can result in
temper or grinding cracks during grinding. Oxidation resulting in decarbuziation
at the surface will cause overheating during grinding and surface temper. Heat-
treat distortion must be minimized so that excessive stock removal is not
required to clean up the part. Often, development programs are required to
determine what type of distortion will occur during heat treating. The parts
can then be cut to take advantage of the movement during heat treating. Use
of quenching dies is another method to limit heat-treat distortion.
Manufacturing Methods 277
278 Gear Drive Systems

HONING

Honing is a finishing process for hardened gears in which an abrasive impreg­

nated plastic helical gear-shaped tool meshes with the work gear in a crossed
axis relationship (Figure 7.10). As the tool and work rotate the honing tool
traverses across the workpiece. Honing is used to improve gear tooth surface
finish, remove nicks and burrs on the teeth, and correct minor errors in the
gear tooth shape. Honing can produce surface finishes as fine as 6 /iin. rms.
It can be applied to external and internal gear teeth and achieve profile
modification and crowning. The amount of stock removed in the honing
process is small: 0.0003 to 0.001 in.

O THER METHODS OF GEAR TOOTH MACHINING

The techniques used to generate and form gear teeth discussed above are not
the only machining processes in use. Gears are also produced by milling, shear
cutting, broaching, stamping, molding, die casting, sintering, and rolling.
These processes either do not produce gears of sufficient strength or accuracy
for the applications covered in this book or are high-production techniques
which require considerable investment in tooling and are not economical for
the relatively small production lots run for this type of gearing.

SHOT PEENING

In this process high-velocity particles are shot at the gear tooth, deforming
the metal surface, in order to produce residual compressive stresses at the
surface. The depressions are very shallow: less than 0.0001 in. Shot peening is
used primarily to induce or increase the compressive stress in the root fillet
area to a level sufficiently high such that the bending stress applied by the
tooth load is offset and fatigue cracks at the surface will not initiate or propa­
gate. Similar beneficial results in pitting resistance are claimed for shot
peening of the tooth flanks; however, some users maintain that shot peening
increases tooth surface roughness and they shot peen the root fillet area only.
This may be accomplished by masking the gear tooth flank during peening
or by shaving or grinding the tooth flanks after peening.
Effectiveness of shot peening depends on the following factors:
Type of shot
Hardness of shot
Uniformity of shot
Manufacturing Methods 279

Velocity of the shot stream

Duration of the treatment
Distance from the nozzle to the workpiece
Angle of impact
In order to develop the right combination of these factors for a given applica­
tion a technique named the Almen strip test is used. The test strip is a stan­
dard piece of spring steel which is shot peened on one side. The resulting
residual surface compressive stresses make the strip bow upward and the
height of the bowed arc is an index of the intensity of the peening. The arc
height can be related to the depth of the residual stress layer and the
magnitude of residual stress.
Steel shot with a hardness of Rc 45 to 55 is commonly used for
through-hardened gears and a shot hardness of Rc 55 to 65 is used for case-
hardened parts. In some cases glass shot is used. In order to penetrate the root
fillet area the shot diameter should be no larger than half of the smallest fillet
radius. The shot quality in terms of spherical shape and uniform size is im­
portant. Sharp-edged particles can damage the tooth surface. The shot stream
should be approximately perpendicular to the surface being blasted. Because
the gear tooth root is contoured, it is usually necessary to blast from more
than one position.

INSPECTION EQUIPMENT

The gear tooth involute profile, lead, spacing, runout, and thickness can be
accurately measured on the bench using standard inspection equipment; how­
ever, specialized machines are available to perform these inspections quickly,
economically, and with the capability of recording the results on a permanent
chart. When all the foregoing elements of a gear tooth are measured individ­
ually, the process is called an analytical inspection. Another technique used is
the functional inspection, where the gear in question is rolled with a master
gear or its mating gear to determine acceptability. When the functional
inspection indicates a problem, an analytical inspection is performed to
determine the cause.
The following paragraphs will describe some of the specialized instru­
ments available for performing analytical and functional inspections of gear
teeth.

Involute Profile Measuring Machine

Figure 7.11 illustrates one type of involute checker. The gear is mounted in
centers and rotates in a timed relationship with a stylus that traverses the
 je
m
PS

Figure 7.11 Involute profile measuring machine. (Courtesy of Fellows Corpora­

tion, Springfield, Vt.)

280
Manufacturing Methods 281

profile in the transverse plane. The deviation, of the profile from a true
involute form is measured and displayed on a dial indicator graduated in
0.0001 in. increments and/or a recording pickup which may have a sensi­
tivity as fine as 0.000020 in. The profile is recorded on a strip chart. To
calibrate the machine an accurate master with known form is inspected.

Figure 7.12 Lead checking machine. (Courtesy o f Fellows Corporation,

Springfield, Vt.)
282 Gear Drive Systems

■ ■ ■ ■ ■ ■

*& *£j

Figure 7.13 True-position spacing checker. (Courtesy of Fellows Corporation,

Springfield, Vt.)
Manufacturing Methods 283

Lead Measuring Machine

A gear tooth lead inspection is an indication of the alignment of the face with
the axis of rotation. In the case of a spur gear the face width should be parallel
to the reference axis. For helical gear teeth the lead check determines if the
tooth has the required helix angle. Figure 7.12 illustrates a lead checking
machine. The gear is mounted between centers and a stylus traverses the face in
an axial direction. For helical gears the gear rotates as the stylus is advanced at
the proper rate to satisfy the lead relationship. If the lead is correct the stylus
will describe a straight line. The deviation from a straight line is displayed on a
dial indicator and/or a strip chart using an electrical pickup. The lead measure­
ment can be made to an accuracy of 0.0001 in. or better.

Tooth Spacing

Two types of tooth spacing checkers are in general use. One measures pitch vari­
ation, the difference between one circular pitch, and the circular pitches immedi­
ately before and after. This type of machine has a fixed finger that acts as a stop
on the pitch line of a tooth. A second finger or stylus senses the position of the
adjacent tooth and actuates a dial indicator or recording pen.
The other type of tooth spacing checker measures index variation or the
true position of a tooth. The gear is mounted on a rotating disk which turns in
increments corresponding to the number of teeth in the gear. A stylus comes in
and contacts each tooth near the pitch line. The first tooth is set at 0.0 and the
deviation of each subsequent tooth from its theoretical position is measured and
recorded. Figure 7.13 illustrates a true-position spacing checker. The two types
of spacing checkers will not give identical readings on the same gear because they
operate on different principles; however, analysis of the findings of either
machine will lead to the same conclusions concerning gear acceptability. On
either machine tooth spacing can be checked to accuracies of 0.0001 in. or
better.

FUNCTIONAL INSPECTIONS

Composite Inspection

Figure 7.14 illustrates a composite gear checking machine. This type of inspec­
tion is sometimes called a red line, in reference to the strip chart that is pro­
duced. The gear to be checked is mated with a master gear which is mounted on
springs such that the center distance between the two gears can vary as the teeth
mesh. The variation in center distance is a function of tooth error and is
measured and recorded. The variation in center distance from tooth to tooth is
a composite measurement of errors in tooth spacing, profile, lead, and surface
 Gear Drive Systems

Figure 7.14 Composite gear checking machine. (Courtesy of Fellows Corpora­

tion, Springfield, Vt.)
Manufacturing Methods 285

1 REVOLUTION OF WORK GEAR

finish. The total composite variation is a measure of the gear runout plus the
local composite variations. Figure 7.15 illustrates a red line chart.
The composite error reading is influenced by errors in the master gear as
well as errors in the part being inspected; therefore, the master must be as
perfect as possible. The composite inspection method is a quick, inexpensive
way of checking production gears. Analysis of the red line chart can differen­
tiate between the various types of errors; however, the composite check is not
sufficiently accurate for measurement of critical gears above AGMA Quality
Class 12 or gears that require accurate profile or lead modifications.

Error in Action

Gear tooth inaccuracies will result in variations of velocity ratio in a pair of

mating gears as the driven gear accelerates and decelerates. Machines are becom­
ing available which measure this variation. One type uses friction disks with
diameters machined to the exact pitch diameters of the gears to be inspected.
One disk is attached to the driving gear and the other disk is mounted on the
same axis but independent of the driven gear. The driven disk operates at a
constant speed and the driven gear angular velocity is compared to it electrically
and the variation recorded. This type of checker measures the action of the
loaded faces of the gear teeth only, unlike the composite checker, where the
286 Gear Drive Systems

mating gears are always in tight mesh with both faces touching. For this reason
it is sometimes referred to as a single flank tooth tester.

Tooth Pattern Inspection

Often gear teeth are rolled together after a marking compound is applied to the
teeth in order to check the contact pattern. This can be done in an inspection
fixture or in the actual gear housing. In an inspection fixture the tooth pattern
will show the amount of contact across the face width and indicate how well
the teeth line up. In the gear housing errors other than tooth alignment come
into play, such as bearing bore parallelism. Sometimes the pattern check is used
to determine a further modification of the gear tooth, which is then remachined
to achieve good contact. When performing this type of check care must be taken
to spread the marking compound very lightly. If the thickness of marking com­
pound is greater than approximately 0.0001 in., the errors that the inspection is
attempting to reveal will be masked by excessive smearing of the compound.

BIBLIOGRAPHY

Dudley, D. W., Gear Handbook, McGraw-Hill, New York, 1962, Chaps. 16 to 23.
Modern Methods of Gear Manufacture, National Broach and Machine Division,
Lear Siegler, Inc., Detroit, 1972.
8
GEARBOX ECONOMICS

There are a multitude of variables affecting gearbox costs which make it difficult
to determine how expensive a unit must be for a given application. The trade-off
between design conservatism and cost is always present, the goal being to achieve
satisfactory performance at minimum price. In each application different criteria
are applied as to what constitutes satisfactory performance. For a relatively inex­
pensive gearbox some period of trouble-free performance after which the unit is
easily replaced may be perfectly satisfactory. In some process industry applica­
tions the cost of downtime incurred if a gearbox must be modified or replaced is
so great that the initial gearbox price in comparison is insignificant. In such a
case satisfactory performance is equated with extremely high reliability over
long periods of time and in order to achieve this reliability, unit costs increase
significantly.
The cost of a gearbox can be divided into three elements:
Material costs
Manufacturing costs
Purchased items such as bearings, seals, lubrication components, and so on.
These three factors are dependent on the gearbox design and configuration
chosen. To illustrate this, let us look at an example of a 500-hp gear unit driven
by a 5000-rpm steam turbine and driving an 1800-rpm generator. This applica­
tion might be satisfied by any of the following designs:
1. Through-hardened cut double helical gearing
2. Through-hardened cut and shaved double helical gearing
3. Single helical, case hardened and ground gearing

287
288 Gear Drive Systems

4. Single helical, case hardened and ground pinion meshing with a through-
hardened cut gear
Design 1 will be the largest unit since the tooth stresses must be low for this rela­
tively inaccurate method of fabrication. Design 3 will be the smallest and there­
fore will have significantly lower material costs. The single helical gearing,
however, requires more machining time since it is cut and ground and also
requires more sophisticated heat treatment. Bearings for design 3 will probably
be more expensive since there are gear thrust loads to deal with. The larger gear
sets of designs 1 and 2 will generate more heat in churning, requiring more
cooling oil flow and therefore a larger pump, cooler, filter, reservoir. These are
only a few examples of many tradeoffs that are made in selecting a gearbox
design.
If one were to solicit competitive quotes from companies making each of
these types of designs it would not be possible to predict which would be least
expensive. If one company’s design were inherently significantly more costly
than the others, the company would not long remain in business. Quite often the
low bid is not a reflection of economical design or manufacture but an indica­
tion of how badly the company wants the particular program.
A cost consideration beyond the gearbox itself is the impact of the unit on
the total system. A smaller gearbox will save costs associated with the base plate.
Also, shipping costs will be less. The gearbox configuration, parallel shaft with
an offset between input and output, or planetary with concentric shafts, will
also affect the system design and influence costs. Planetary gears lend themselves
to close coupling of the gearbox to the driver or driven equipment, with the
potential of saving coupling and shafting costs.
From the discussion above it can be seen that if one wanted to determine
the most cost effective gearbox for a new application and were free to choose
any manufacturing and heat treating methods, a detailed design study and cost
analysis of the numerous potential design solutions would be required to arrive
at a sensible conclusion. More often companies buy gearboxes complete from a
manufacturer with existing facilities or have their own machine tools and
facilities; therefore, the design chosen reflects the type of gearbox the company
has made in the past. The cost question then focuses on what materials to use
and what degree of quality to require in the gearbox components.

GEAR QUALITY

The types of gears being discussed in this book range in American Gear Manufac­
turers Association (AGMA) Quality Class from about 8 for the less accurate end
to 14 for the most accurate. A Quality Class 8 gear might be used in low-speed
mining equipment, while a Quality Class 14 gear might be used in a sophisticated
Gearbox Economics 289

very high speed compressor drive. Automotive gears are typically Quality Class
10 to 12. The quality achieved is to some extent determined by the finishing
process; however, the various processes overlap in terms of quality produced. An
approximate comparison follows:

AGMA quality range Process

8-12 Hobbing and shaping
9-13 Shaving
10-14 and up Grinding

Figure 8.1 presents an approximate relationship between AGMA quality

class and cost. The numbers are relative, with a Quality 14 gear costing approxi­
mately 8 times as much as a Quality Class 8 gear. The figure shows that at the
high-quality end of the spectrum a small increase in quality requires a large
expenditure. Conversely, at the low end decreased quality offers little savings.
AGMA standards give recommendations for quality classes suitable for various
applications. As experience is gained in a particular service it is possible to
ascertain more accurately which tooth tolerances are important and how closely
they must be held. It would then be appropriate to buy gears to a specification
reflecting that experience rather than to a general quality specification. To be

RELATIVE COST

Figure 8.1 Approximate relationship between AGMA quality class and cost.
290 Gear Drive Systems

certain that the quality specified is reflected in the gear tooth geometry, a user
should request inspection documentation. These might take the form of red line,
involute, lead, or tooth spacing charts. Measurements over balls, bore parallelism,
center distance measurements, and so on, might also be requested to be docu­
mented. There may be additional costs for this documentation. The need for and
extent of inspection documentation is a matter of judgment, but if AGMA
quality class requirements of 12 or above are specified, it is recommended that
100% documentation be requested. For Quality Class 14 the user and manufac­
turer should agree beforehand on how the inspections will be made and the
results interpreted. For quality classes lower than 12 it is good practice to docu­
ment prototype units so that if field problems occur in subsequent production
gearboxes, a baseline of quality has been established.
So far, our discussion of quality has been confined to gear tooth geometry.
If it is determined that high quality is required in the gear teeth, undoubtedly
other components in the unit must also be specified to equivalent quality levels.
In fact, failures associated with bearings, the dynamic system, and the lubrica­
tion system are more common than gear tooth problems; therefore, complete
specification of all components in the unit in addition to the gears is necessary.

MATERIAL COSTS

Gear steels are bought on a dollars per pound basis either as bar stock or in a
forged state. Up to a diameter of approximately 6 in. it is cheaper to buy bar
stock than forgings. Above 6 in. bar stock becomes harder to procure and
forgings have an economic advantage, particularly since forgings can be worked
closer to the final shape than bar stock and therefore save machining costs.
Taking a low-alloy steel such as Society of Automotive Engineers (SAE)
1040 as a base, the cost of high-alloy steels used in gears such as SAE 4340 or
9310 may be three or four times as high. Of course, gears made of high-alloy
steel will usually be smaller than those of low-alloy steel, so the cost of material
is somewhat offset by the smaller quantity. If a requirement for vacuum-melt
steel rather than air-melt steel is specified, the price may be doubled. Over the
years the quality of gear steels has deteriorated and it is not uncommon to
encounter impurities or voids in air-melt steels. These discrepancies are some­
times found only after several machining operations have been completed; there­
fore, the higher cost of vacuum-melt steels may be offset by a reduction of
scrapped work in process.
The question often arises: Should the gear casing be a welded steel fabri­
cation or a casting? From a cost point of view, for a single unit a fabrication is
cheaper. This is because a pattern must be machined prior to pouring the casting,
and developing the pattern is a costly and time-consuming procedure. As the
Gearbox Economics 291

quantity of gearboxes required increases, pattern costs can be amortized over

the production run and castings become economically attractive. The casting
can be poured close to the final shape, and extensive machining and welding can
be eliminated. For a typical gear unit, the break-even point may be as low as
three gearboxes. Castings may be made of cast iron or steel. A steel casting costs
approximately 30% more than a cast iron part. On occasion, gear housings are
made of aluminum to save weight. Aluminum is twice as expensive as steel on a
dollar per pound basis, but because the housing weight will be much lower, it
is not clear which is more expensive. Aluminum housings, however, usually have
to incorporate steel bearing liners and inserts for threads, which increase cost.

EFFECT OF QUANTITY ON COSTS

When manufacturing gearbox components, the time to set up the equipment far
exceeds the time required to machine the parts. Therefore, it is far more expen­
sive to perform a machining operation on one part than on many parts. For
instance, it may take 8 hr to set up a gear grinder and 2 hr to grind a part.
Assuming a cost of $X/hr, one part would cost $10(X), whereas four parts
would cost $4(X) each [(8/4 + 2)(X )].
If a single special unit is required for an application, design and tooling
costs must be considered. These costs will probably exceed the manufacturing
expenses. It is therefore worthwhile to attempt to use a standard gear unit or
one that has been manufactured for a different application, even if it requires
compromising the system design to some extent.

BIBLIOGRAPHY

Hamilton, J. M., Are You Paying Too Much for Gears, Machine Design, October
1972, pp. 144-150.
Kron, H. O., Optimum Design of Parallel Shaft Gearing, ASME Paper 72-PTG-17,
October 1972.
The Cost of Gear Accuracy, Design Engineering, January 1981, pp. 49-52.
AGMA Gear Handbook 390.01, Vol. 1, Gear Classification, Materials and
Measuring Methods for Unassembled Gears, American Gear Manufacturers
Association, Arlington, Va., 1973.
9
PLANETARY GEAR TRAINS

A planetary gear train is one in which the power is transmitted through two or
more load paths rather than the single load path of a simple gear mesh. Figure
9.1 illustrates the components of the simplest type planetary, a sun gear, planet
gears, a ring gear, and the planet carrier. When the carrier rotates about the
center of the system a point on a planet gear not only rotates about the axis of
the planet gear but also about the center of the system and this type of drive is
called epicyclic.
In the United States, planetary gears are widely utilized in automotive
automatic transmissions and in aerospace drives such as turbine engine reduction
gears or helicopter transmissions but are not used extensively in industrial appli­
cations. There is a trend toward increasing utilization of planetary gears because
they offer the following advantages:
1. Because they share load between several meshes, planetaries are more
compact than parallel shaft drives and offer significant envelope and weight
savings. An example illustrates this point:
Let us take a 1000-hp electric motor at 1800 rpm driving a compressor at
7200 rpm. The required American Gear Manufacturers Association
(AGMA) service factor is 1.6.
A hardened and ground, helical, parallel shaft gear set would have
the following dimensions:
Pinion pitch diameter: 4.0 in.
Gear pitch diameter: 16.0 in.
Face width: 4.0
Gearbox envelope: 24 X 12 X 12 in.
Gearbox weight: 650 lb

293
294 Gear Drive Systems

The equivalent planetary gear set with a stationary ring gear and rotating carrier
would have the following dimensions:
Sun pitch diameter: 3.5 in.
Ring pitch diameter: 10.5 in.
Planet pitch diameter: 3.5 in.
Face width: 3.0 in.
Gearbox envelope: 15 in. diameter X 10 in. length
Gearbox weight: 250 lb.
Figure 9.2 illustrates the gearbox size comparison.
2. In addition to achieving minimum weight and envelope, the relatively
smaller and stiffer components which result from the use of planetary gearing
lead to reduced noise and vibration and increased efficiency. Noise and effi­
ciency are strongly dependent on the speed of the components. In the example
above, the parallel shaft pitch line velocity is 7540 fpm compared to the plane­
tary gear pitch line velocity of 4948 fpm, a significant difference. The lower
planetary pitch line velocity is a result of the smaller sun gear and the epicyclic
Planetary Gear Trains 295

PARALLEL SHAFTCONFIGURATION PLANETARY CONFIGURATION

action of the gear mesh. Later in this chapter the method of calculating epicyclic
speeds is derived.
3. The input and output shaft axes in planetary gearing are concentric.
This can lead to space savings in some installations by allowing the driving and
driven equipment to be in-line. The coaxial feature is very important in auto­
matic transmissions since it makes rapid speed changes possible without the
necessity of taking gears out of mesh. The rotating components are controlled
by the use of clutches and brakes to achieve speed changes.
4. The resultant radial forces on the input and output shafts of planetary
gearboxes are zero since the arrangement cancels out all radial forces. In effect,
the planetary gearbox transmits only torque. This simplifies bearing design and
in some applications allows close coupling of the gearbox to the driving or driven
equipment.

ANALYSIS OF PLANETARY GEAR ARRANGEMENTS

Planetary gears can be arranged in a multitude of configurations to achieve

specific ratios and power splits. This section presents a method of calculating
planetary speed ratios and power flows. The method is applied to several
examples and by following the logic the reader can generate equations describing
the mechanics of any planetary configuration. The following nomenclature will
be used throughout.
Symbols
R = gear pitch radius, in.
W = gear angular velocity, rad/sec
296 Gear Drive Systems

WT = tangential load, lb
n = gear rpm
Subscripts
s = sun gear
p = planet gear
r = ring gear
c = carrier
1 = primary stage(high-speed stage)
2 = secondary stage(low-speed stage)
For example, the symbol RSl would be the pitch radius of the primary-stage sun
gear.

Simple Single-Stage Planetary

Figure 9.1 shows the general case where the sun gear, ring gear, and planet
carrier are all free to rotate. Directions of rotation are assumed to be as shown in
the figure. With all members rotating this is termed a differential system and
there are potentially six different ways to connect prime movers and driven
equipment:

Inputs Outputs
Sun Carrier, ring
Carrier Sun, ring
Ring Carrier, sun
Sun, ring Carrier
Carrier, sun Ring
Ring, carrier Sun

The rotating elements in each of the arrangements above have a distinct speed
and torque relationship. In order to define the angular velocity of all three
elements in the planetary gear train, the angular velocity of two elements must
be specified. Let us derive the generalized speed relationships of the gear train
in Figure 9.1.
1. A point on the pitch diameter of the sun gear has a tangential velocity of
WsRs-
2. The point on the sun gear pitch diameter is meshing with a point on the
planet gear pitch diameter which has a tangential velocity made up of two
components, WpRp + WCRS.
Planetary Gear Trains 297

PLA

B
" WpR, * W c R s
'WpRp

Figure 9.3 Idealized planetary system.

Figure 9.3 shows the planetary system idealized as friction disks with a point A
on the sun disk meshing with a point B on the planet disk and also shows their
respective tangential velocities. Because the points are meshing they must have
the same tangential velocities; therefore,
(9.1)

(9 .2 )

Note that at point C on the planet gear (Figure 9.3) the components of the
tangential velocity are in opposite directions and subtractive, whereas at point
B they are in the same direction and additive. The tangential velocity of the
planet carrier is zero at the center of the system and equals WCR at any radius R.
Combining Eqs. (9.1) and (9.2), the general speed ratio equation is
WSRS = WrRr + WC(RS + Rr) (9.3)

Let us apply this equation to the example shown earlier in this chapter:

In this case the ring gear was stationary; therefore,

Wr = 0
298 Gear Drive Systems

W, R, + Rr
Ratio = — = —-------- = 4
Wc R„
/Rs \
In a simple gear train the rpm of the planet gear would be rpm sun • I — 1;

however, because of the epicyclic action the planet rpm is reduced and can be
calculated from either Eq. (9.1) or (9.2) with
rpm sun = 7200
and
rpm carrier = 1800
the planet rpm = 5400 and the pitch line velocity of the sun planet mesh is
pLV = 5400(3.5)7r = 494g f
12
The rolling velocity of the sun with the planet is 5400(3.5/3.5) = 5400 rpm and
its absolute velocity is 7200 rpm.
If, in the system shown in Figure 9.1, the carrier is held stationary, Wc = 0
and Eq. (9.3) becomes

Ws _ Ri

Wr Rs
In this case, where there is no epicyclic action, the speed ratio is simply the ratio
of the sun and ring gear radii. This configuration is sometimes called a star
system.
If the sun gear is held stationary, Ws = 0 and Eq. (9.3) becomes

Wr _ Rr + RS
Wc ' " Rr
The negative sign indicates that the sense of rotation of one of the components
shown in Figure 9.1 was chosen opposite to the direction in which it actually
rotates. In this case, if the ring gear is driving, the carrier would rotate in the
same direction as the ring gear rather than the direction shown in Figure 9.1
(viewed from the sun gear end).
In the equations above rpm can be used in place of angular velocity and
number of teeth in place of pitch radius. Quite often, planetary gear equations
are presented in this manner.
If all components of a planetary gear set are free to rotate, the speed ratios
are dependent on the power split between the components. Figure 9.4 shows the
tangential tooth loads at the sun and ring gear mesh points. The torque and
horsepower of each component are as follows:
Planetary Gear Trains 299
RING

Figure 9.4 Planetary tangential tooth loads.

Torque
Component (in.-lb) Horsepower

WTRsns
Sun WxR,
63,025

Rs + Rr WX(RS + Rr)nc
Carrier 2Wt s 1
1 2 63,025
WxR tnr
Ring wTRr
63,025

It should be noted that the tangential load WT used in the table is the summa­
tion of the tangential loads at each planet gear mesh.
Let us assume that in the previous example the 1800 rpm carrier with a
1000 hp input drives a compressor which is attached to the sun gear and also
drives an oil pump which is attached to the ring gear. The oil pump absorbs
300 Gear Drive Systems

50 hp. The three horsepower equations in the previous table can be solved for
the three unknowns WT, Ns, and Rr. Using the carrier horsepower equation with
R s = 1.75, R r = 5.25, and nc = 1800 the tangential load WT = 5002 lb. Knowing
WT , the sun and ring horsepower equations yield
ns = 6840 rpm
nr = 120 rpm
It should be noted that in this discussion no attempt was made to account for
the power loss due to friction in the train.
The arrangement shown in Figure 9.1 is capable of producing gear ratios
in a range of approximately 2:1 to 10:1. The speed ratio derived earlier in this
section is Rr/Rs for the stationary carrier, rotating ring gear case. In this situa­
tion the input and output directions of rotation are reversed. With a stationary
ring gear and rotating carrier, the speed ratio is 1 + (Rj/Rg) and the input and
output rotation directions are the same. At low ratios the sun gear diameter
approaches the ring gear diameter and therefore planet gear diameters become
small and impractical. As one strives for the higher ratios in a single-stage
planetary, the sun gear becomes smaller and is stress limited. Also, at ratios
above approximately 10 the ring gear becomes large and more economical
arrangements are available.

Compound Planetary

For a ratio of approximately 10:1 to 16:1, the compound planetary gear set
(Figure 9.5) provides a practical and economic configuration. The speed reduc­
tion equations for this design follow:
II

WSRS + WCRS
*Xt

WrRr ^P2^P2 - WcRr

ii

WP1
o*

Figure 9.5 Compound planetary gear.

Planetary Gear Trains 301

Combining these equations, we have

In the case of a stationary ring gear and rotating carrier the input and output
direction of rotation are the same and the speed ratio is

In the case of a stationary carrier and rotating ring gear the input and output
directions of rotation are opposite and the speed ratio is

For low ratios in the range 3:1 a reverted type of compound planetary is used as
shown in Figure 9.6. The speed ratio for this train is simply

S2 1VP 1
RSiR p2
In this case the input and output directions of rotation are reversed.

Multistage Planetaries

By combining stages of planetary gearing in various arrangements, large ratios

can be achieved. For instance, let us take two stages of simple planetary gearing
as shown in Figure 9.1 and combine them in two ways as shown in Figure 9.7A
and 9.7B. Figure 9.7A presents a drive where the primary-stage carrier drives the
secondary-stage sun gear. The reduction ratio of the primary stage is

w<si
Ri
w,ci
The reduction ratio of the second stage with a carrier output is

and the total reduction is

302
Figure 9.6 Reverted compound planetary. (Courtesy of AVCO, Lycoming Division, Stratford, Conn.)
Planetary Gear Trains 303

Figure 9.7 Two-stage planetary gearing.

The input and output rotation are in the same direction. Figure 9.7B shows a
ring gear output from the second stage and in this case the second-stage reduc­
tion ratio is

R - " k
2 w r2 r S2

and the total reduction ratio is

R = R t R2 = (l + —\ —
\ RS1/ R S2

The input and output directions of rotation are opposite in sense.

304 Gear Drive Systems

Split Power Transmissions

An interesting planetary arrangement which results in a very compact design is

the split power transmission. Figure 9.8 illustrates one version which is most
useful in a ratio range of approximately 15:1. As shown in the figure the
primary planetary stage is driven by the sun gear and has both a rotating carrier
and ring gear. The primary ring gear drives the sun gear of the second stage
which through the stationary second-stage planets drives the second-stage ring
gear. Both the second-stage ring gear and the primary stage carrier are con­
nected to the output shaft, thus enabling the power split. Part of the input
power is transmitted directly to the output shaft through the primary stage
carrier, thereby bypassing the second stage of gearing. Following are the speed
and power equations describing the split power transmission:

WS1 Rs, = WP 1 R P1 + WC1 R S1

Wr i RI l = Wp i Rri - WClRri
Planetary Gear Trains 305

WS2Rs2 = Wp2Rp2 = 2

wvr2 = w ci
where the subscripts Sl, Cl, p i , and ri refer to the primary-stage sun, carrier,
planet, and ring and the subscripts S2, p2, and r2 refer to the second-stage sun,
planet, and ring gears.
Combining the equations above we arrive at the expression for the speed
ratio:

R = Wsa = Wj, = l + /Rn \ A / R r2 Rn

Wei Wr2 \ R Si / \ R S2 R si
To determine the horsepower split between the primary and secondary stage, the
following expressions are derived:

Trp = WTRs i nsi

HFin
63,025
where nSl is the primary sun gear rpm.

(Rsi + Rr i )nc
HP primary carrier = 2WT
2(63,025)
where nc is the carrier or output rpm. The percentage of power going out the
carrier is
HP primary carrier (RSl + Rr i )nc 1 / Rri \
HP^ R si^ sj R \ + R ^7/
The exact percentage of total power transmitted directly by the primary carrier
is dependent on the primary- and secondary-stage ratio split but is approxi­
mately one-third of the total. The remaining two-thirds of the power is trans­
mitted through the secondary stage. The power split-up serves to reduce
secondary-stage loading and increases gear efficiency. The disadvantage of the
split power system is that it is mechanically complex.
Figure 9.9 shows a different split power variation which is used to achieve
ratios in the range 24:1 to 55:1. The application here is a wheelmotor which
combines a hydraulic motor and the planetary gearing as a single unit in the
wheel hub. Wheelmotors are used in mobile construction and off-the-road
equipment where a centrally located pump drives the hydraulic motors at each
wheel.
Figure 9.10 illustrates a wheelmotor casing. In the arrangement of Figure
9.9 the hydraulic motor drives through a sun gear. The output members are the
primary- and secondary-stage ring gears which are connected to the wheel rim.
The power split up is as follows: Part of the input goes directly from the primary
306 Gear Drive Systems

Figure 9.9 Wheelmotor transmission. Key: 1, casing; 2, vehicle chasis; 3, road

brake; 4, hydraulic motor; 5, drive shaft; 6, first planetary stage; 7, second plan­
etary stage; 8, axle journal; 9, hub; 10, rim; 11, twin tires; 12, roller bearing.
(Courtesy of American Lohmann Corporation, Hillside, N.J.)

sun to the primary ring gear. The remainder goes directly from the primary
carrier to the secondary sun gear and then to the secondary ring gear.

Power Feedback Systems

When working with complicated gear systems it is possible to arrange the gears in
such a manner that the power transmitted by some components is greater than
the input power. Although these arrangements generally achieve unusually large
ratios, the component size must be sufficient to handle the recirculating power
and this disadvantage may offset the large ratio obtained.
If it is not recognized that a system has recirculating power, failures may
occur. Figure 9.11 illustrates a feedback system. The input is to the primary-
stage sun gear, which is connected to the secondary-stage sun gear. The output is
the secondary-stage carrier. Following are the speed equations for this system:
Planetary Gear Trains 307

Figure 9.10 Wheelmotor casing. (Courtesy of American Lohmann Corporation,

Hillside, N.J.)
308 Gear Drive Systems

WS1RS1 = Wr i Rri = WpjRpj

Wr2Rr2 ~ ^P2^P2 “ ^C2^T2

Ws2RS2 = WP2RP2 +
W
vvs 2 = vvsi
W
Wr 2 = Y
WVri
Combining the equations above the speed ratio is
WS1 _ 1 + r I2/ r S2
R
WC2 ^ R siR r2
R n RS2
Let us assume the following dimensions:
RS1 = 1 in.
RS2 = 1.05 in.
Rri = 8.3 in.
Rr2 = 8.25 in.
The reduction ratio R = 166:1. Let us determine the horsepower transmitted:
R„_ + Rr nc
hp out = 2WT c --------- ------ --------
F TS2 2 63,025

where
WXs 2 = secondary-stage sun tangential load, lb
nC2 = carrier rpm

u A WTS2Rs2Ns2
hp secondary sun = ----------------
63,025
where NS2 is the secondary sun in rpm.
hp secondary sun / RS2 x
R = 18.74
hp out \R S2 + Rj 2 ^
The power circulating in the gear train is 18.74 times the output power.

Planetary Gear Design Considerations

Although the basic gear tooth design of planetary gear configurations is no

different from parallel shaft gearing, there are several points that must be
Planetary Gear Trains 309

considered when rating planetary gears and defining the detail geometry. This
section discusses load sharing, assembly, and choice of numbers of teeth and
numbers of planets.

Load Sharing

The main advantage of planetary gearing, of course, is transmittal of load

through two or more parallel paths. Ideally, the load should be shared equally
by each planet gear, but due to manufacturing errors this is never the case. Some
of the factors affecting load sharing are:
Inaccurate carrier bore location
Inaccurate gear tooth spacing
Variation in planet tooth thickness
Eccentricities between sun, ring, and carrier
Figure 9.12A shows the load distribution of a three-planet system. The situation
is such that in order for the sun gear to be in static equilibrium the planets must
share load equally. If the sun gear is free to move and one planet is overloaded,
the sun gear will shift to equalize the planet loads. Figure 9.12B shows static
equilibrium; however, the load is not equally shared between planets.
In general, there are three methods of attacking the problem of load
sharing between planets:
1. A completely rigid system which relies on precise component tolerances
2. A system with flexibility built in by floating one or more of the members
3. Systems that rely on mechanical means to adjust the planets to provide load
equalization
The most straightforward approach is a combination of 1 and 2, where toler­
ances are closely held and the sun or ring or both are flexible to adjust to load
maldistributions. This technique is used in high-speed planetaries such as those in
gas turbine engines. Strain gage studies have shown that load sharing within 10%
can be practically achieved.
AGMA Standard 420.04, Practice for Enclosed Speed Reducers or
Increasers Using Spur, Helical, Herringbone and Spiral Bevel Gears, states the
following:
To compensate for unequal loading of multiple planet pinions, the total
power capacity of all pinions should be the calculated capacity of one
pinion plus a maximum of 0.9 times the calculated capacity for each
additional pinion. For two planet pinions multiply the calculated capacity
of one pinion by 1.9; for three pinions multiply by 2.8. If a load balancing
device is used which insures equal loading of all planets, the capacity
modifying factors need not be used.
310 Gear Drive Systems

Figure 9.12 Planetary load distribution. A. Three planet system—load must be

equally shared if sun is free to float. B. Four planet system in static equilibrium
with unequal load distribution.
Planetary Gear Trains 311

For any specific application the gear designer must determine what degree of
load sharing is anticipated and rate the gearset accordingly. This estimate may be
based on analysis, test, or experience, but it is important to understand that
perfectly equal load sharing will not be achieved.

STOECKfCHT DESIGN

A planetary configuration, extensively used in Europe for many years, is com­

monly called the Stoeckicht design, named after the inventor. Figure 9.13 illus­
trates this double helical configuration, which features flexibility not only in the
sun gear but also in the ring gears, which are connected to the housing by a series
of splines. The purpose of the splines is to achieve load equalization not only
between the planets but also between each of the two helical gears. This system
has been very successful in the past since the many degrees of freedom tend to
compensate for tooth errors. There are disadvantages to the Stoeckicht design:
1. Use of through-hardened double helical gearing does not achieve as small an
envelope as hardened and ground single helical gearing.
2. The design is mechanically complex.
3. The design does not lend itself to compounding to achieve higher ratios.
312 Gear Drive Systems

Planetary Gear Assembly

When choosing the numbers of teeth in planetary gears, it must be understood
that not every combination of gear teeth can be assembled. For instance, Figure
9.14 shows a three-planet set with a six-tooth sun gear and an 18-tooth ring gear.
In this instance it is obvious that planet gears can be placed into mesh at the
0°, 120°, and 240° positions and there is no assembly problem. However, if the
ring gear had 19 teeth the gearset would not be assembleable. For a simple
planet set (sun, planet, ring) to be capable of being assembled, the following
equation must be satisfied:
Planetary Gear Trains 313

on

where
H = a whole number (integer)
Ns = number of sun gear teeth
Nr = number of ring gear teeth
K = number of planets
To understand this equation, refer to Figure 9.15. Assume a three-planet system
with a planet gear located at the 0° position as shown. In order to assemble a
planet at the 120° position, the centerline of a tooth or tooth space of the sun
gear must arrive at the 120° point at the same time as the centerline of a tooth
or tooth space of the ring gear. Let A be the angle between the center of that
sun gear tooth or space and the 120° line and S be the angle between that ring
gear tooth or space and the 120° mark. The following equation can be written
for A:
360 360FL
A = ----- - ------- -
K Ns
where
K = number of planets
Hs = whole number of sun gear teeth
Ns = total number of sun gear teeth
Note that the equation is general and can be applied to systems with any number
of planets.
314 Gear Drive Systems

A similar equation can be written for the ring gear angle S:

„ /360 TT\ 360
Nr 7 K
where
Hr = whole number of ring gear teeth
Nr = total number of ring gear teeth
Line ZZ on the planet (Figure 9.15) passes through either two tooth centers or
tooth spaces or one tooth center and one tooth space, depending on whether the
planet has an even or odd number of teeth. The planet rotates through an angle
B at the same time the sun rotates through angle A and the ring gear through
an£le S. The relationship between these angles is
AN, SNr
B = — = —-
np np
where Np is the total number of planet teeth. Substituting for A and S yields
360HS\ Ns /360Hr
Ns / Np \ Nr
which simplifies to the assembly equation
N, + Nr
H = —--------
K
The assembly equation for a compound planet system is more complicated and
can be understood by considering Figure 9.16. Assume that points M and N are
registry teeth on the compound planet. By this is meant that the centerline of a
tooth on the primary planet is aligned with the centerline of a secondary planet
tooth. The purpose of the registry marks is to enable all compound planets to be
manufactured identically.
As was shown in the analysis of the simple planet system, the center of a
sun gear space rotated an angle A from the 120° mark must arrive at the 120°
mark at the same time as the center of a ring gear space, which is shown rotated
an angle S from the 120° tooth mark.
360 360H,
A = s
K Ns

360H. 360
S = ----- ----------
Nr K
Planetary Gear Trains 315

The primary planet rotates through an angle B at the same time the sun rotates
through A:
AN,
B = ---- -
N Pi
where Np x is the total number of primary planet teeth. The secondary planet
rotates through an angle G at the same time the ring gear rotates through S.
GNr
S =
Np2
where Np2 is the total number of secondary planet teeth.
The relationships between the registry teeth and the meshing teeth on the
primary and secondary planets are:

j = 360HP,
N P,

0 . 36Q NPa
316 Gear Drive Systems

where
Hp x = whole number of primary planet teeth
Hp2 = whole number of secondary planet teeth
From Figure 9.16,
I = G + O - B
Substituting for I, G, 0 , and B, we have
360Hpi _ Nr / 360Hr 360\ 360Hp2 Ns ^360 360HS
Npi N p 2\ \N r K
^ / Np 2 Np i l \K Ns /
and simplifying yields
Nr + NSND2/ND1 ,
—---------g- £ g _ Pl + (H - Hs) - S I = Hr + H„
K Npi r P2
In order to solve this equation the term on the left,
N
i^r + N sNp 2'/N p l
K
is calculated and whole numbers are substituted for (Hp t - Hs). If the left side
of the equation works out to be a whole number, the gearset can be assembled.
In a compound planetary each of the planets must be identical not only
for assembly purposes but also for proper load sharing. The registry teeth are
used to satisfy this requirement and a typical allowable tolerance for the varia­
tion of the centerline of the primary planet tooth with the centerline of the
secondary tooth is ±0.001 in. Figure 9.17 is a photograph of a compound plane­
tary gear. Compound planets can be either of two-piece or one-piece construc­
tion. The processing must be very precise to achieve the registry dimension. In
addition to the registry, the tooth thickness and stock removal must be closely
controlled. The sequence of operations for a two-piece hardened and ground
compound planet are as follows:
1. Cut primary and secondary planet teeth
2. Heat treat
3. Finish grind secondary planet
4. Press primary onto secondary using a fixture to align registry teeth
5. Grind primary planet to final tooth size and registry dimension
The reverted compound planetary gearset (Figure 9.6) with an input and output
sun gear has the following assembly equation:
Planetary Gear Trains 317

’•-a*-

Figure 9.17 Compound planet being balanced.

where
N0 = total number of low-speed sun gear teeth
H0 = whole number of low-speed sun gear teeth
It should be noted that it may be physically possible to assemble a gearset that
does not satisfy the assembly equations or has mistimed registry markings. If
there is sufficient backlash it may be possible to force the components into
mesh; however, the gears at some point in the rotation will tend to bind and
will have poor load sharing.

Choice of Tooth Numbers

There are three considerations in choosing the numbers of teeth of a planetary in

addition to the basic problem of achieving the correct ratio:
1. Assembly considerations as discussed in the preceding section
2. Choosing hunting teeth
3. Achieving sequential mesh rather than simultaneous mesh
318 Gear Drive Systems

By hunting teeth it is meant that the numbers of teeth in a gear mesh be such
that each tooth in the pinion at some point in time mesh with each tooth in the
gear. For instance with a 20-tooth sun and 40-tooth planet each sun gear tooth
would always mesh with the same 2 planet teeth. By changing the number of
planet teeth to 41, a hunting tooth mesh is achieved.
The advantage of having a hunting tooth is that the gearset has more of a
tendency to wear in. For instance, if one tooth has a high spot and meshes with
all the mating gear teeth, there is a better chance that the error will wear in and
not cause permanent damage than if the discrepant tooth kept mating with the
same two meshing gear teeth.
Sequential mesh occurs in a sun-planet mesh when the number of teeth in
the sun is not divisible by the number of planets. When the number of teeth in
the sun is divisible by the numbers of planets the mesh is called simultaneous. In
this case each sun-planet mesh is occurring at the same point on the gear at the
same time. For instance, when the planet at the 0° point is in mesh at its pitch
line, the 120° and 240° planets in a three-planet system are also being loaded at
the pitch line. This situation should be avoided, if possible, since the load
impulse the simultaneous mesh gives to the mechanical system is stronger than
that of the sequential mesh. The sequential mesh, therefore, is less likely to
generate harmful vibrations.
The mesh frequency of a simple gearset is the number of teeth times the
speed. If a planetary gear has sequential meshing, the mesh frequency will be the
number of planets times the number of teeth times the speed. The higher
sequential mesh frequency is less likely to excite any natural frequencies in the
operating equipment.
Although having hunting teeth and sequential mesh is desirable, it is not
always achieved and many gearsets operate successfully without fulfilling these
conditions. The most obvious way of determining the number of planet teeth is
to subtract the number of sun teeth from the number of ring teeth and divide
by 2. Let us take an example:
Ns = number of sun gear teeth = 20
Nr = number of ring gear teeth =100
CD = center distance, inches = 10
The planet gear geometry is
Planetary Gear Trains 319

where DP is the diametral pitch.

Np
PDn = —- = 13.3333 in.
p DP
What if the number of ring gear teeth minus the number of sun gear teeth
is not even? Let us assume the following case:

Ns = 20

Nr = 103
CD = 10

NrI ■ Ns
NP = = 41.5

If we make the number of planet teeth 41 there is no problem in achieving a

practical gear set, but the operating conditions are different at the sun-planet
mesh from the planet-ring mesh. At the sun-planet mesh:
Ns + Nd
DP, „ = —-------2 = 3.05
sp 2(CD)
The planet pitch diameter is

PDP<S-P> ■ ^ * 134426
^ r s-p
At the planet-ring mesh:

Nr - N0
DP„ r = — 2- = 3.10
p'r 2(CD)

PDp(p-r) = AJr p_r = 1 3 '2258

If the pressure angle 0 is chosen at 22.5° at the sun-planet mesh, the pressure
angle at the planet-ring mesh is calculated as follows:
COS0S_P • PDp(s_p) = cos0p.r • PDp(p_r)
an d 0 p.r = 20.1117°. The lead of the planet gear is
7T • PD
L =
tan 0
If the helix angle 0 is chosen as 10° at the sun-planet mesh, the helix angle at
the planet-ring mesh is as follows:
320 Gear Drive Systems

tan tf ( } = tan tf( } • p(p-lj-

FDP(P-s)
and i//p.r = 9.8419°. The fact that the operating conditions at the sun-planet
mesh are different from those at the planet-ring mesh is advantageous. The
sun-planet mesh is critical in terms of compressive stress and flash temperature
rise; therefore, it is desirable to have a large pressure angle which reduces both
these parameters. Because the ring-planet mesh is internal, compressive stress and
flash temperature rise are not usually a problem and the lower pressure angle can
be tolerated. In fact, a lower pressure angle results in lower separating forces on
the ring gear and this is desirable to reduce ring distortion and stress under load.
Even in the case where the number of ring gear teeth minus the number of sun
gear teeth is an even number, it can be advantageous to drop one or two teeth
from the planet in order to achieve a larger pressure angle at the sun-planet mesh
pitch diameter and a smaller pressure angle at the planet-ring mesh pitch
diameter.
The question may be asked: How many planets should one use in a
particular planetary set? There is a physical limit to the number of planets that
can fit into any specific application. Figure 9.18 plots the maximum number of
planet gears which can be assembled into a planet system versus the ratio of the
sun gear diameter divided by the planet gear diameter. As the reduction ratio
becomes smaller, the sun gear becomes larger and more planets can be fit in. As
shown earlier, in a three-planet system with a floating sun gear, the load between
the planets tends to be equalized. As the number of planets is increased, load
sharing may suffer and the full benefit of the additional planets may not be
realized. The vast majority of planetary gears use three planets; however, systems
with 4, 5, 7, and more planets have been successfully developed.
A point that must be considered when planet gears are idlers meshing with
both the sun and ring gear is that the planet teeth experience complete reversal
of stress. Allowable bending stresses must be multiplied by 0.7 to compensate
for this loading condition.

Planetary Bearing Loads

Quite often the critical components from a life point of view in a planetary
system are the bearings. Because the planet gears are loaded both at the sun and
ring gear meshing points (Figure 9.4), the planet bearings must react twice the
tooth loads.
In an epicyclic system where the planet carrier rotates about the center of
the system, the centrifugal force on the planet gears must be reacted by the
planet bearings. The centrifugal force on a planet gear is calculated as follows:
Fc = mrW2
Planetary Gear Trains 321

CO
hHI
z
<
CL
u.
O
a
uj
CD
2
3
Z

2
3
S
X
<
s

SUN GEAR DIA. / PLANET GEAR DSA.

Figure 9.18 Maximum number of planets that can be assembled.

where
Fc = centrifugal force, lb
m = planet mass, lb-sec2/in.
r = radius to planet center, in.
W = carrier angular velocity, rad/sec
Let us work through an example of a helical epicyclic planetary gear set
and calculate the planet bearing loads.
PDS = sun gear pitch diameter, in. = 3.5
PDp = planet gear pitch diameter, in. = 3.5
HP = transmitted horsepower = 1000
= sun rpm = 7200
322 Gear Drive Systems
RESULTANT BEARING RESULTANT BEARING
RADIAL LOAD RADIAL LOAD
R=1667 + *70 3= 18 09 # R=1667 + *90 8= 18 98 #

|s = 6 0 7 #
703# I | 908#
| T ° 2^ ® W t -1 6 6 7 # J*

j -1 6 6 7 # ® ”
“2 9 4 #

r IS = 6 0 7 #

Figure 9.19 Planet bearing loading.

nc = carrier rpm = 1800

0 = helix angle, deg =10
0 = pressure angle, deg = 20
Figure 9.19 illustrates the planet configuration and loading. Assume that the
planet weighs 5 lb. The centrifugal force is

Fc = — ( 3 . 5 ) ’ « 161,1b
‘ 316 1 60 /
The tangential load per planet with a three-planet system is

6 3 fl2 5 0 9 ® (±\ I . 16671b

1 7,200 \3 .5 / 3
The separating load at each mesh is
S = W tan 0 = 607 lb
Planetary Gear Trains 323

The thrust load at each mesh is

T = Wx tan ^ = 294 lb
For the planet to be in static equilibrium the bearing loads are as shown in
Figure 9.19. It can be seen that the centrifugal load is significant. The limiting
factor in an epicyclic system quite often is the carrier speed, which, if excessive,
results in overloading of the planet bearings due to centrifugal force.

Planetary Gear Economics

At first glance it would appear that a planetary gearset must be more expensive
than a parallel shaft configuration. There is the added complication of a carrier
and two or more additional planet gears plus more bearings. Ring gears are
usually expensive items.
In fact, the reduced size of planetary components offsets the cost of addi­
tional parts and the determination of which design is less costly is not so
obvious. An important factor is the quantity to be manufactured and as the
quantity of units increases, the savings in material and advantages of handling
and machining smaller components begin to outweigh the disadvantages of
mechanical complexity.
Also, when high ratios are required (over 15:1) planetaries tend to be more
economical, even in relatively small quantities.
In very high horsepower units, the components in parallel shaft boxes tend
to become so large that there may be an economic advantage to planetaries even
in the ratio range 2:1 to 15:1.
In any case, the technical advantages of planetary gearing are sufficient
such that this type of gearing should be considered, especially since the
economic penalty may be small or nonexistent.
10
GEARBOX INSTALLATION:
MOUNTING, ALIGNMENT, COUPLINGS

Proper installation of the geared system is essential to achieve good performance.

The gearbox must be rigidly connected to the foundation, which must also be
rigid and have a flat mounting surface. If the foundation or base plate structure
is incorrectly designed or constructed vibration, shaft misalignment, bearing
damage, and even shaft or housing breakage can result.
Most gearboxes are foot mounted. A flange on the bottom of the unit is
doweled and bolted to a base. The doweling is important to ensure that the
gearbox does not move during operation. This is a four-point mounting arrange­
ment, and since three points define a plane it is difficult to install the unit such
that all four points lie in the same plane. Two major reasons why the mounting
points on the base plate are usually not coplanar are:
1. The steelwork warps as a result of poor welding, grouting, or concrete work.
2. The use of multiple steel beams which are not coplanar in the base plate.
When mounting the unit on multiple steel beams a base plate which extends
under the entire gearbox and is at least as thick as the gearbox base should be
used. Both the gearbox and the base plate should be rigidly bolted to the steel
supports.
Shims can be used to bring all mounting surfaces into the same plane.
Care must be taken that the shims form a solid tight pack when the bolts are
tightened. Prior to final tightening, the shims should be inspected for rust, folds,
wrinkles, burrs, tool marks, and dirt. Correct grouting is of great importance
when bases are supported on concrete. Figure 10.1 is a checklist to ensure
proper grouting.

325
326 Gear Drive Systems

Proper grouting checklist

1. Use non-shrink grout, as confirmed 7. Place grout continuously, as

by test data per ASTM C-827. Volume quickly as possible, and work from
change after hardening should be the one side or end across to the other side
measurement used. or end, to avoid trapping air in con­
2. Make sure baseplate design permits fined spaces.
a complete fill, with adequate pour 8. Ensure that the surrounding tem­
openings and inspection holes. perature and the chemical additives
3. If epoxy grout is to be used, ask for allow ample working time for proper
evidence that long-term cold flow or placement before setup (at least 45
“compression creep" will not occur. minutes for a full foundation).
4. Avoid metallic aggregate grouts 9. Allow enough curing time before
where either severe temperature swings placing the drive in operation; depend­
or corrosive surroundings may exist. ing on the type of grout, this could be
5. Keep the concrete foundation moist as long as 7 days.
for 24 to 48 hours prior to grouting. 10. When the drive is in service, watch
6. Be sure all oil, grease, or dirt is for signs of deterioration of grout.
removed from any surface to be con­
tacted by grout.

One result o f ignoring the checklist above.

Figure 10.1 Grouting procedures. (From Ref. 1.)

Gearbox Installation 327

Stability of the foundation is extremely important. Once installed, the base

must not deform. One major cause of difficulty is thermal expansion, which can
be due either to ambient conditions such as partial sunshine unevenly heating
the structure or proximity to hot operating equipment. Of course, the structure
must initially be designed with sufficient rigidity to withstand all operating
forces without distortions.
It is possible to encounter a resonance condition where the natural frequency
of the complete system assembly, including the base plate, coincides with an
operating frequency. This type of structural resonance can be corrected in
several ways:
1. The natural frequency of the base can be increased by making it more rigid
by adding stiffeners or gussets.
2. The natural frequency of the base can be lowered by removing material from
the base.
3. The mounting system can be made more elastic to dampen or lower the fre­
quency. An example of this is the use of spring washers at the attachment
points.
4. The mass of the rotating system can be changed to modify its natural fre­
quencies.
When handling the unit at installation, care must be taken not to stress parts
which are not meant to support the gearbox weight. Gearboxes should be lifted
only by the means provided by the manufacturers, such as lifting holes in the
casing.

COUPLINGS AND SYSTEM ALIGNMENT

To connect the driving and driven equipment to the gearbox, input and output
shaft couplings are used. It would not be practical to align the centerlines of the
equipment exactly; therefore, the coupling must have some degree of flexibility
to accommodate misalignment. Even if it were possible to perfectly align the
equipment at some given operating point, because load, speed, or ambient condi­
tions vary, the alignment will change. Figure 10.2 illustrates the alignment condi­
tions that the equipment shaft ends may be in. The coupling must accommodate
these conditions while transmitting torque, yet limit the forces on machine com­
ponents such as shafts and bearings that result from misalignment. If the
coupling had no flexibility, the misalignment would have to be accommodated
by bending of the shafting, which would apply loads to the bearings. The maxi­
mum amount of misalignment couplings can accommodate must be satisfied at
assembly with alignment procedures that will be described below. It is impor­
tant, however, to attempt to minimize the misalignment to the lowest practical
values since coupling life is strongly dependent on how well the system is aligned.
328 Gear Drive Systems

A LIG N M E N T

- /
PARALLEL OFFSET MISALIGNMENT

SYMMETRICAL ANGULAR MISALIGNMENT

Figure 10.2 Shaft alignment conditions. (From Ref. 2, courtesy of American

Gear Manufacturers Association, Arlington, Va.)

Figure 10.3 illustrates the most widely used coupling types. With gear couplings,
misalignment is accommodated by sliding of the spline teeth. The sliding results
in wear, and excessive wear is the major cause of failure of this type of coupling.
In order to limit wear, gear couplings are lubricated, usually with grease. The
maximum sliding velocity is [3]

v = ? e 2ir ‘ rpm
s 2 60
where
Vs = sliding velocity, ips
6 = misalignment angle, in./in. (rad)
D = pitch diameter, in.
Gearbox Installation 329

DISC COUPLING

DIAPHRAGM COUPLING

Figure 10.3 Coupling types.

Reference 3 gives the following guidance for allowable sliding velocities:

Less than 1.4 ips: optimum condition
3.0 ips: normal operation
5.0 ips: maximum allowed
Gear couplings can accommodate axial motion by allowing the male splines to
float axially within the female splines as shown in Figure 10.3.
 330 Gear Drive Systems

RUBBER ELEMENTS

------------------------- ---------

■ ■ i ■ s p

Figure 10.4 Elastic coupling. (Courtesy of American Lohmann Corporation,

Hillside, N.J.)

Disk and diaphragm couplings accommodate misalignment and axial motion

by flexing of the elastic elements. These couplings generally fail in fatigue and
the fatigue life is dependent on misalignment, coupling speed, and the applied
steady-state loads due to torque transference. Disk and diaphragm couplings
require no lubrication.
Elastic couplings which incorporate rubber elements bonded to a steel
backing are sometimes used with diesel drives to damp torsional vibrations.
Figure 10.4 illustrates an elastic coupling with rubber elements which dampen
vibration and accommodate misalignment.
The maximum amount of angular misalignment flexible couplings camaccom-
modate is in the order of %°. The maximum allowable offset is dependent on the
distance from pivot to pivot. For instance, if the distance from the center of one
spline to the center of the other spline on the gear coupling shown in Figure
10.3 is 10.0 in. and %° or 0.0044 in ./in. maximum angular misalignment is
possible, the maximum inches of parallel offset are
0.0044(10.0) = 0.044 in.
The values above are the maximum flexible couplings can accommodate;
however, operation at these limits will shorten coupling life and may be detri­
mental to the system. During operation misalignment should not exceed
Gearbox Installation 331

approximately one-fifth of the foregoing values and the system should be care­
fully aligned prior to startup to achieve this.

ALIGNMENT PROCEDURE

An outline of the steps required to align two pieces of equipment connected by

a coupling is as follows:
1. Roughly set up the machines visually and with crude measurements.
2. Accurately align the shafts in the cold condition using precise measurements.
3. Operate the equipment and accurately measure the alignment in the hot
condition.
4. Calculate the cold alignment required to achieve alignment in the hot condi­
tion. For instance, if it is determined that a motor shaft rises 0.005 in. with
respect to the gearbox shaft during operation, the motor shaft should be
positioned 0.005 in. below the gearbox shaft during cold alignment.
5. Realign in the calculated cold condition, operate the system, and recheck
alignment in the hot condition.
The alignment procedure begins with the coupling hubs mounted on their
respective shafts. The equipment is placed in position with the proper axial gap
X, as shown on Figure 10.5. In order to roughly align the shafts, crude measure­
ments can be made with a straightedge and calipers. The straightedge checked in

Figure 10.5 Rough alignment of coupling halves.

332 Gear Drive Systems

Figure 10.6 Checking for angular misalignment.

two planes will give an indication of the parallel offset. Caliper measurements at
four points will give an indication of the angular misalignment.
There are several methods available to accurately align the shaft ends. One
method, using a dial indicator, will be described. The first step, as shown in
Figure 10.6 is to check for angular misalignment. A dial indicator base is
mounted securely on the right-hand hub and the dial indicator stem is placed
against a face on the left-hand hub. The connected shafts are rotated several
times and the dial indicator checked to ensure ample movement in either
direction. It is useful to use a mirror to observe the dial indicator gage as the
shaft is rotated. The point at which a minimum reading is registered is found and
at this point the dial indicator gage is set at zero. When the shaft is rotated 180°
from this point the dial indicator reading will be the total angular misalignment.
As an example, refer to Figure 10.6. If the minimum reading occurred at the
bottom position, when the coupling is rotated one-half turn and the dial
indicator is at the top position as pictured, the indicator will read the dimension
X, which means that the distance between the coupling faces is X inches greater
at the top than at the bottom. In order to align the faces, the equipment must
be shimmed to narrow the distance between the faces at the top to X/2 inches.
The distance between the faces at the bottom will increase X/2 inches and the
faces will be square.
Gearbox Installation 333

Now that the angular misalignment is corrected, parallel offset must be

measured. Figure 10.7 shows the dial indicator base securely fastened to the
right-hand coupling hub with the stem in contact with a smooth outside
diameter on the left-hand hub. Again the shafts are rotated several turns, making
sure the indicator has travel in both directions and the point where the reading is
minimum is found. The dial indicator is set at zero at this point and the shaft
rotated 180°. In this position the indicator will read twice the amount of parallel
offset. For instance, in Figure 10.7 if the minimum indicator reading was at the
bottom position, when the coupling is rotated one-half turn and the indicator
stem is at the top position, as pictured, if the reading is 0.050 in. the parallel
offset of the center lines of the shaft ends is 0.025 in. with the left-hand shaft
higher than the right-hand shaft. The equipment can now be shimmed vertically
by 0.025 in. to bring the coupling into alignment.
After correcting for parallel offset the axial spacing and angular measure­
ment should be rechecked to make sure that they were not disturbed. If the
bracket holding the dial indicator is not rigid and allows the indicator to sag, an
error will be introduced into the alignment readings. Readings in the horizontal
plane will be little affected, but in the vertical plane, sag will increase the reading
when the indicator is on top and decrease the reading when the indicator is on
the bottom. One way to measure if the dial indicator is sagging is to mount the
breacket on an accurately machined cylinder. The top reading, T (Figure 10.8),
is greater than the nominal reading by the amount of sag and the bottom
reading, B, is less than the nominal reading by the amount of sag. The sag is

The parallel offset determined by the method shown in Figure 10.7 must be
corrected for the sag. This can be done by lowering the centerline of the
measured shaft by the amount of sag. For instance, if the centerline of the left-
hand shaft is measured to be 0.025 in. higher than the right-hand shaft and the
334 Gear Drive Systems

Figure 10.8 Measuring dial indicator bracket sag.

sag is measured to be 0.005 in., the actual parallel offset in the vertical plane is
0.020.in.
In Figures 10.6 and 10.7 both shafts are rotated when making the angular
misalignment and parallel offset measurements. It would be possible to rotate
only the shaft holding the dial indicator bracket and traverse the face and
outside diameter of the other shaft while it is stationary. In this case any outside
diameter eccentricity or face runout will be a source of error in the alignment
reading. This problem is averted by rotating the measured surface together with
the indicator.
Dial indicator methods of measuring misalignment have the disadvantage of
not being able to monitor machine condition while operating. There are optical
and proximity probe systems which can be used to monitor alignment continu­
ously when an application warrants the expense of such instrumentation [3].

REFERENCES

1. Nailen, R. N., Trouble Free Drives Are Based on Firm Foundations, Power
Transmission Design, May 1981, pp. 38-42.
2. AGMA Standard 510.02, Nomenclature for Flexible Couplings, American
Gear Manufacturers Association, Arlington, Va., August 1969.
3. Cox, J. L. and Wilde, L. G., Alignment of Turbomachinery, Sawyers Turbo-
Maintenance Handbook, Vol. Ill, Turbomachinery International Publica­
tions, Norwalk, Conn., 1980.

BIBLIOGRAPHY

Flexible Coupling Two Step Dial Indicator Method, Rexnord Inc., Coupling
Division, Warren, Pa.
11
GEAR UNIT OPERATION:
TESTING, STARTUP, CONDITION
MONITORING

In this chapter three phases of gearbox operation are discussed: testing, initial
field startup, and condition monitoring.

TESTING

During the procurement process the gearbox manufacturer and the user must
agree on the type of testing the completed unit will be subjected to prior to
acceptance. The test program can be as simple as turning the shafting by hand
to verify free operation of the internal components or as complicated as full-
scale operation of an instrumented gearbox, on the actual application, through
a predetermined test schedule.
Several factors must be considered when determining the degree of test
program complexity:
1. Cost of testing
2. Confidence in the gearbox design
3. Consequences of gearbox failure
Clearly, a proven gear design, operating in an environment where downtime
does not incur a large cost penalty, would not warrant extensive testing. On the
other hand, if a gearbox is a critical component in a complex system the cost
of testing may be slight compared to the loss in case of failure. Where human
safety considerations are involved, testing may be required to limit legal liability.
New designs or extrapolations of existing designs warrant sufficient testing to
verify the analytical and manufacturing procedures.

335
336 Gear Drive Systems

Test programs can be extremely valuable in identifying problem areas prior

to field operation. Design, manufacturing, and assembly errors can be identified
and corrected prior to a costly catastrophic failure. In order to gain the most
information, the test plan and instrumentation scheme should be carefully
designed and acceptance limits set prior to operation. Data should be taken over
the range of speeds, loads, and operating environments anticipated in the
application. To identify potential problem areas quickly, the test program
should include overspeed and overload operation. Care must be taken not to go
too far and develop failure modes that would not occur in the actual application,
but overspeeds to 110% and overloads to 125% should be within the capability
of the unit.
Some of the common problems that can be identified in initial testing are:
1. Excessive gearbox heat generation. This is most commonly caused by oil
churning and can be corrected by improving scavenging, reducing oil flow or
changing oil type.
2. Improper gear pattern. May be caused by gear tooth errors, bore misalign­
ment, or deflections. Can be corrected by tooth modifications or possibly
bearing bore relocation.
3. Overheating o f gears or bearings. Usually caused by insufficient lubrication.
Can be corrected by increasing oil flow, retargetting oil jets, or changing oil
type.
4. Excessive noise or vibration. Caused by unbalance, tooth errors, assembly
errors, or operating at critical frequencies.
5. Oil leakage. Caused by misassembly of static or dynamic seals or pressuriza-
tion of gearbox cavity.

Spin Tests

In many cases a full-speed, light-load test is considered sufficient to qualify a

gear unit as acceptable. A typical test program might be:
1. Operate the gearbox at maximum continuous speed until bearing and lubri­
cation oil temperature has stabilized.
2. Increase the speed to 110% of maximum continuous speed and run for a
minimum of 15 min.
3. Reduce speed to maximum continuous and run for 4 hr at a minimum.
The following measurements should be made during the acceptance test:
Oil inlet temperature
Scavenge oil temperature
Oil feed pressure
Oil flow
Shaft speed
Gear Unit Operation 337

M O D E L ___________________ S / N ________

DATE_________ OPERATOR _______________ ENGINEER

RUN TIME ..5PEEG..... OIL TEMP.. °F OsF pM ^ ur I OIL FLOW VIBRATION

NO. OF DAY RPM 1 2 3 PSIG GPM LEVEL

Figure 11.1 Typical test log sheet.

Other parameters that can be measured during gear testing are:

Vibration
Shaft excursion
Noise
Bearing temperatures
A detailed test log should be kept making entries of each measurement at regular
intervals such as every 15 min. Figure 11.1 shows a typical test log.
After completion of the mechanical running test, the gear unit should be
opened for a visual inspection. Tooth meshes should be inspected for surface
damage and proper tooth contact. All bearings and journals should be inspected
for signs of surface damage or overheating.
With high-speed gear drives it is not uncommon to conduct the full-speed
acceptance test driving the gearbox through a low-speed shaft. This is done if a
high-speed prime mover is not available. In this case, the gears are contacting on
their normally unloaded faces since the gearbox is being driven backwards. Such
a test can still be useful to determine proper operation of the lubrication system
and correct alignment of the gear shafts. Also, any gross machining or assembly
errors can be identified.
The full-speed “light” load test is widely used since full-load testing at a
gear vendor’s plant is costly and sometimes a prime mover of sufficient capacity
is not available.

Load Testing

Relatively low power gearboxes (up to possibly 200 hp) are load tested with
power absorption devices loading the output shaft or shafts. Power absorption
338 Gear Drive Systems

✓HYDRAULIC MOTOR RADIATOR

ORIFICE

HYDRAULIC PUMP

Figure 11.2 Power absorption test setups.

devices generally convert the power into heat and include water brakes, dyna­
mometers, generators with load banks, hydraulic motors driving pumps, and so
on. Prime movers include electric, hydraulic, or gasoline motors. Figure 11.2
illustrates some power absorption test setups.
For gear units transmitting high horsepower, power absorption testing is
expensive and sometimes impractical. Quite often a suitable prime mover or
power absorption device is not available to the gear manufacturer. Also, the
power required to conduct tests becomes a major expense. To test large units
Gear Unit Operation 339

TORQUE IS APPLIED
BY RELATIVE TWIST
OF COUPLING HALVES

SECTION AA

.TEST UNIT /TORQUE .SLAVE UNIT

at full speed and power, regenerative power techniques are used. Some other
terms for this method of testing are recirculatory power, four-square rig, or
back-to-back testing.
Figure 11.3 illustrates the simplest type of regenerative power test rig.
Two identical gear sets are used. The high-speed shafts are coupled together, as
are the low-speed shafts. Torque is applied to either the high- or low-speed shaft
340 Gear Drive Systems
10000 HP SLAVE UNIT 10000 HP TEST UNIT

Figure 11.4 Back-to-back gear rig.

by twisting one coupling half flange with respect to the other and the torque is
locked in by clamping the coupling halves together. The assembly is then rotated
at the desired speed by the prime mover, which need only develop enough power
to overcome the friction losses in the system.
To illustrate this point, Figure 11.4 shows two gear units set up in a back-
to-back arrangement. If each unit transmits 10,000 hp and is 98% efficient the
loss in the gearboxes will be 400 hp. Therefore, the prime mover for this
10,000-hp can be sized at 400 hp plus some margin for losses other than in the
gear units.
With the arrangement shown in Figure 11.3, any gear tooth load desired
can be developed. One gear pair is designated as the test set and the other the
slave set. On the test set the torque and rotation are applied in the same sense
as on the actual application. The slave set gear teeth, while loaded on the proper
tooth faces, are rotating in the opposite sense. Another way to look at it is that
in the test set if the pinion is driving the gear, the opposite is true in the slave
set, with the gear driving the pinion. A review of the directions of rotation and
torque shown in Figure 11.3 will illustrate this point.
When designing a back-to-back test thought must be given to the bearing
and lubrication system operating conditions of the slave set since the opposite
sense of rotation may require incorporation of some modifications. For instance,
the oil feed groove in a journal bearing may have to be relocated.
The four-square type of rig gives reliable results concerning gear unit
deflections under load and can be used to establish tooth modifications to
Gear Unit Operation 341

Figure 115 Helical gear loading regenerative rig.

improve load distribution. It is also useful in establishing efficiency values and

exploring lubrication system problems. Long-term endurance testing can be
accomplished at low cost.
The test setup shown in Figure 11.3 has some disadvantages:
1. The load is applied at zero speed when there is no oil film generated at the
gear tooth and bearing interfaces. This may cause surface distress of the
mating components. Also, it is impossible to simulate actual operating condi­
tion of most applications where load increases with speed. Starting torque
requirements on the prime mover are high because full load is applied at
zero speed.
2. As gears and bearings wear the locked-in load will gradually decrease.
3. The load may vary with operating temperature as components distort. This
problem can be resolved by retorquing the coupling when rig temperature
stabilizes.
342 Gear Drive Systems

Figure 11.6 Hydraulic rotary torque actuation rig.

From these three points it can be seen that it would be advantageous to be able
to apply and control the load while the rig is rotating and this can be accom­
plished by various means. One method shown in Figure 11.5 makes use of the
fact that axial movement of a helical gear will cause rotation of the shafts. As
the gear translates, all backlash is taken out of the system and torque is gener­
ated. The torque is proportional to the oil pressure applied at the load piston
and therefore can be controlled at all speeds. Another method of loading a rig
during operation is the rotary torque actuator shown in Figure 11.6. Pressure oil
is fed to the actuator and the relative rotation of the vanes generates torque
proportional to the oil pressure.
Parallel shaft units lend themselves to a four-square type of rig arrange­
ment. Figure 11.7 shows a planetary gearbox rig where the high-speed shafts of
the test and slave gearboxes are connected and operating concentrically inside
the low-speed shafts. In order to apply load to the system, one of the stationary
planet carriers is mounted on a loading fixture and turned to introduce a torque
Gear Unit Operation 343

Figure 11.7 Planetary recirculating power rig.

into the system. Figure 11.8 illustrates such a test setup with pneumatic pistons
supplying a couple to turn the carrier.
Another type of recirculatory power test setup is shown in Figure 11.9. In
this case only one test gearbox is required. The electric motor, through the
gearbox, drives a generator, which in turn generates electricity to drive the
motor. Additional power is required to offset energy losses in the system. The
same technique can be used with a hydraulic motor and pump.

Static Tests
It is possible to apply full load to a stationary gearbox and determine tooth
patterns and housing rigidity, as shown in Figure 11.10. The tooth patterns will
344 Gear Drive Systems

show up if bluing or a paste compound is applied to one gear. If a paste com­

pound is applied, care must be taken to minimize the thickness since paste thick­
nesses greater than 0.0002 in. will mask the existence of misalignment between
the gear teeth.
If equipment is available to conduct a low-speed, full-torque rolling test,
this type of operation can be specified. For instance, a 10,000-hp, 20,000-rpm
shaft can be operated at 200 rpm with a 100-hp prime mover and develop the
same torque, 31,512 in.-lb. This type of test is useful to demonstrate tooth
contact, load-carrying capability, and housing rigidity, but yields no information
concerning parameters related to speed, such as temperature, vibration, and
noise.
Gear Unit Operation 345

Figure 11.9 Regenerative electric motor drive rig.

SHAFT FIXED TO GROUND

Figure 11.10 Static test rig.

346 Gear Drive Systems

Instrumentation

At a minimum during a gear test torque, speed, oil temperatures, and pressures
should be monitored. Also important are vibration and oil flow.
Prior to a test an instrumentation plan should be formulated which
includes the type of instrumentation, location, and accuracy. Typical accuracy
tolerances of common measurements are as follows:

Speed ±0.5%
Torque ±2%
Power ±2%
Pressure ±2%
Temperature ±2%
Flow rate ±5%
Vibration frequency ±2%
Vibration amplitude (sine) ±10%

Instruments should be calibrated periodically and whenever possible an instru­

ment reading should be verified by an independent measurement to determine
if it is operating properly. For instance, if a flowmeter is used to measure oil
flow, at some point a known volume of oil should be collected and the time
recorded to check the flowmeter accuracy. Instrumentation failure is common
and must be anticipated; therefore, redundant readings and backup systems
must be planned for.
The most convenient way to measure speed is with a toothed wheel incor­
porated in the shaft system. An electromagnetic transducer is activated by the
teeth passing by and the rpm are digitally displayed via an electronic counter.
Care must be taken to set the proper gap between the tooth tips and the trans­
ducer. Also, runout of the gear can give erroneous readings. The counter gear
usually has 60 teeth, but any number of teeth can be used if the electronics
are set up to compensate. Other ways to measure speed are with a hand-held
tachometer placed in contact with a shaft, and with strobe lights. When con­
ducting a test, speed should always be measured in two different ways to ensure
that a mistake is not made. At initial startup the correct direction of rotation
should be verified.
Torque can be measured by monitoring shaft twist with strain gages. An
indirect way of determining torque is by measuring the forces at the gearbox
mounting pads (Figure 11.11). The housing torque is either the sum or dif­
ference of the input and output torques depending on whether the input and
output shafts are rotating in the same or opposite directions. In Figure 11.11
assume the following:
Gear Unit Operation 347

Figure 11.11 Gearbox housing torque measurement.

F = 10001b
D = 10 in.
therefore, the housing torque is 10,000 in.-lb. If the gear box ratio is 2:1, the
input torque Tj is one-half the output torque TQ, and if the input and output
shafts rotate in the same direction:
T h s g = To -T i = H T0

and the output torque is twice the housing torque, or 20,000 in.-lb.
Temperature, pressure, and vibration measurements are covered later in
the section “Condition Monitoring.”
348 Gear Drive Systems

Reporting

The effort expended in conducting a test may be wasted if the test procedures
and results are not carefully documented. The test items should be documented
with part and serial numbers. Test setups should be described with photographs
and sketches. Test results should be presented completely in the form of tables
and graphs. If failures occur, they should be extensively documented as well as
the modification incorporated to resolve the problem. The final report may be
referred to years after the test program has been completed and the memory
dimmed. Following is a general outline for a final report:
Objective
Summary
Test equipment
Test item
Test plan
Results
Discussion
Conclusions and recommendations

SPECIAL TESTS

Sound Testing

Measurement of the sound level of gearboxes is becoming increasingly important

as government standards concerning equipment noise generation become more
stringent. In addition to noise testing a brief discussion of sound fundamentals
is presented in this section.
As a gearbox vibrates, a pressure oscillation in the surrounding medium
(usually air) is generated. The transmission of the pressure vibration is called a
sound wave and as it travels through the medium it can be detected by some
form of receiver such as a microphone or the human ear. The sound pressure
level of a gearbox is conventionally specified in decibels (dB) at a given distance
from the gearbox. The sound pressure level Lp is the ratio of the pressure of the
sound being measured to a reference pressure:

L = 20 logjo
Fo
where
p = sound pressure being measured, N/m2
pQ = reference pressure, N/m2
cNote : psi X 6893 = N /m 2 .)
 Gear Unit Operation 349

The reference pressure is conventionally taken as 20 pN/m2 (20 X 10"6 N/m2),

which is approximately the threshold of normal hearing at a frequency of 1000
Hz. As an example, if the sound pressure 2 ft from a gearbox is measured as
0.0010 psi (6.893 N/m2), the sound pressure referred to 20 /iN/m2 is

A, B, AND C ELECTRICAL WEIGHTING NETWORKS FOR THE SOUND-LEVEL METER

These numbers assume a flat, diffuse-field response for the sound-level m eter and m icrophone.

A-WEIGHTING B-WEIGHTING C-WEIGHTING A-WEIGHTING B-WEIGHTING C-WEIGHTING

FREQUENCY RELATIVE RELATIVE RELATIVE FREQUENCY RELATIVE RELATIVE RELATIVE
Hz RESPONSE, RESPONSE, RESPONSE, Hz RESPONSE, RESPONSE, RESPONSE,
dB dB dB dB dB dB

10 -7 0 .4 - 3 8 .2 - 14.3 500 -3 .2 -0 .3 0
12.5 -6 3 .4 -3 3 .2 - 1 1 .2 630 -1 .9 - 0 .1 0
16 -5 6 .7 - 2 8 .5 -8 .5 8 00 - 0 .8 0 0
20 - 5 0 .5 - 2 4 .2 -6 .2 1 ,000 0 0 0
25 -4 4 .7 -2 0 .4 -4 .4 1,250 -i 0.6 0 0
3 1 .5 -3 9 .4 - 17.1 -3 .0 1 ,6 00 -f 1.0 0 -0 .1
40 - 3 4 .6 - 14.2 -2 .0 2 ,0 00 -f 1-2 -0 .1 - 0 .2
50 - 3 0 .2 - 11.6 - 1.3 2 ,5 0 0 + 1.3 -0 .2 -0 .3
63 - 2 6 .2 - 9 .3 -0 .8 3 ,1 5 0 + 1.2 -0 .4 -0 .5
80 - 2 2 .5 - 7 .4 -0 .5 4 ,0 0 0 + 1.0 - 0 .7 - 0 .8
100 - 19.1 - 5 .6 - 0 .3 5 ,0 00 + 0 .5 - 1 .2 - 1 .3
125 - 16.1 - 4 .2 - 0 .2 6 ,3 0 0 - 0 .1 - 1 .9 - 2 .0
160 - 13.4 -3 .0 - 0 .1 8 ,0 0 0 - 1 .1 - 2 .9 -3 .0
200 - 10.9 - 2 .0 0 10,0 00 -2 .5 - 4 .3 - 4 .4
250 -8 .6 -1 .3 0 1 2 ,5 00 -4 .3 - 6 .1 - 6 .2
315 - 6 .6 -0 .8 0 1 6 ,0 00 -6 .6 - 8 .4 - 8 .5
400 - 4 .8 - 0 .5 0 2 0 ,0 0 0 - 9 .3 - 1 1 .1 - 1 1 .2

FREQUENCY RESPONSES
(dB)
RESPONSE
RELATIVE

FREQUENCY (Hz)

Figure 11.12 Sound meter frequency response. (From Ref. 1.)

350 Gear Drive Systems

6.89 N /m 2
Lp - 20 log10 = 110.7 dB
20 X 10"6 N/m2
The sound-level meters that are used to measure decibels have a frequency
response which is better than the human ear. In other words, at specific fre­
quencies the meter will record sound a human being would not hear. To com­
pensate for this fact, various filters have been incorporated into sound meters
so that the measurements will approximate what the ear would record. As shown
in Figure 11.12, three weighing scales, A, B, and C, have been established. The A
scale matches the human ear’s response at sound levels below 55 dB, the B scale
at levels between 55 and 85 dB, and the C scale above 85 dB. The A scale is most
commonly specified due to its use by OSHA for measurements up to 115 dB.
Therefore, in the example above, the gearbox sound level measured by a meter
set on the A scale would be defined as 110.7 dBA.
In order to analyze the source of gearbox noise*sound meter readings can
be filtered to register only a limited range of frequencies. Frequently, octave
and 1/3-octave bands are specified, as defined in Table 11.1.
Figure 11.13 presents the results of an octave band analysis of a two-stage
parallel shaft gearbox with the high-speed mesh operating at 18,000 rpm. The
lower curve is a measure of the background noise, including the prime mover
decoupled from the gearbox. When even finer analysis of noise data is required,
filters with bandwidths down to 2 Hz are used. A real-time analyzer can be
employed to look at all frequencies simultaneously.
American Gear Manufacturers Association (AGMA) Standard 295.04 [2]
defines the instrumentation and procedures to be used for sound measurement
of high-speed helical and herringbone gear drives and presents typical maximum
sound levels as shown in Figure 11.14.
Sound test speed and load conditions, according to the standard, are to be
agreed upon by the manufacturer and purchaser. The microphone is to be
located perpendicular to the center of the gear unit’s vertical surface, but not
less than 1 ft above the floor or plate. The distance between the gear unit’s
vertical surface and microphone is to be the normal working distance of the
closest employee or as in the following table:

Gearbox center Microphone distance

distance (in.) (ft)
4 and below 3
4 -1 5 5
15 and over 10
Gear Unit Operation 351

Table 11.1 Continuous Octave and One-Third Octave Frequency Bands

Frequency (Hz)
Octave One-third octave
Lower Upper Lower Upper
band band band band
Band limit Center limit limit Center limit
12 11 16 22 14.1 16 17.8
13 17.8 20 22.4
14 22.4 25 28.2
15 22 31.5 44 28.2 31.5 35.5
16 35.5 40 44.7
17 44.7 50 56.2
18 44 63 88 56.2 63 70.8
19 70.8 80 89.1
20 89.1 100 112
21 88 125 177 112 125 141
22 141 160 178
23 178 200 224
24 177 250 355 224 250 282
25 282 315 355
26 355 400 447
27 355 500 710 447 500 562
28 562 630 708
29 708 800 891
30 710 1,000 1,420 891 1,000 1,122
31 1,122 1,250 1,413
32 1,413 1,600 1,778
33 1,420 2,000 2,840 1,778 2,000 2,239
34 2,239 2,500 2,818
35 2,818 3,150 3,548
36 2,840 4,000 5,680 3,548 4,000 4,467
37 4,467 5,000 5,623
38 5,623 6,300 7,079
39 5,680 8,000 11,360 7,079 8,000 8,913
40 8,913 10,000 11,220
41 11,220 12,500 14,130
42 11,360 16,000 22,720 14,130 16,000 17,780
43 17,780 20,000 22,390
Source: Ref. 1.
352 Gear Drive Systems

CO
Q.

Figure 11.13 Octave band analysis (end location at 3 ft and 14,200 rpm).

Figure 11.14 Typical maximum sound pressure levels versus high-speed mesh
pitch line velocity. Note: In case of multireduction or increasing gear sets within
one housing use the pitch line velocity of the highest speed set. (From Ref. 2.)
 Gear Unit Operation 353

Figure 1 1 . 1 5 Sound test microphone position. Key: L, length of gear unit;

H, height of gear unit; W, width of gear unit; D, distance of microphone perpen­
dicular to unit, as specified in standard for size, h, height of microphone perpen­
dicular to floor (H/2); d, distance of microphone from corner of unit (1/2) or
(W/2). *Note: Load is optional for factory testing. (From Ref. 3.).

Figure 11.15 shows the sound test microphone position. The sound level is to
measured with and without the gear unit operating so as to correct the unit
rating for the ambient sound pressure level. Corrections are presented in
Table 11.2.

Table 11.2 Corrections for Ambient Sound Pressure Levels

Difference between gear unit Correction to be subtracted from
and ambient sound pressure levels gear unit sound pressure level
(dB re 20juN/m2) (dB re 20juN/m2)
3 or less 3
4 and 5 2
6-9 1
10 or greater 0
Source: Ref. 2.
354 Gear Drive Systems

The overall sound level of a mechanical system is made up of the sound

levels of the components therein. For instance, the components of a generator
drive system when measured individually may exhibit the following noise levels:
Prime mover: 88 dBA
Gearbox: 82 dBA
Generator: 95 dBA
The following expression is used to combine the component noise levels to
analytically define an overall noise level:
N
Lp = 101ogI() £ 1001ai
i = 1

where
N = number of components
aA = sound level of each component
In the example,
Lp = 101og10 (108-8 + 108-2 + 109'5)
= 9 6 dBA
It can be seen that in order to reduce the noise level of this system, the major
contributor, the generator, must be worked on.
Noise generation is influenced by the design and manufacture of the gear­
box components. The ideal situation in a gear mesh would be to transmit power
with no change in the angular velocity of the gear shaft. In such a perfect gear
mesh there would be no accelerations or decelerations of the gear shafts, which
provide the energy for vibration and noise generation. Of course, there are no
perfect gear meshes, and errors in tooth spacing, profile, or runout will always
be present, resulting in accelerations and decelerations of the gear shafts and
noise.
Design of gear teeth can reduce noise levels. For instance, increasing
overlap or changing from spur to helical gearing will reduce noise. In these cases
the changes result in smoother load distribution from tooth to tooth as they go
through the mesh. Changes in design to reduce noise, however, might adversely
affect other parameters. For instance, increasing contact ratio by going to a finer
pitch might compromise bending strength. Other design changes that tend to
reduce noise are:
Reduced pitch line velocity
Proper profile and lead modifications
Lower pressure angles
Gear Unit Operation 355

It must be realized that detail gear tooth geometry changes can only account for
changes in sound level of up to approximately 4 dBA. Also, no matter how
optimum the design, the quality of the gearing will determine the sound level.
Table 11.3 lists the common sources of sound in a gear drive system.
Sound can be transmitted through the air or be structure borne. Struc­
ture-borne noise may travel through the support structures and radiate at some
point other than its source. Because of this the noise measurements near a
gearbox in an operating system may be quite different from those measured
during a bench test. Also, when connected to the prime mover and load, the
torsional response of the system may be different from the bench test and result
in different noise measurements.
In many cases when it is determined that the noise generation of a system
is excessive yet not an indication of some malfunction, the most practical
method of noise reduction is the use of an acoustical enclosure. The effective­
ness of such an enclosure is very dependent on eliminating all openings. Such an
enclosure will also affect the ambient temperature and housing heat dissipation
ability of the gearbox and means must be provided to cool the unit properly.

Efficiency Testing

Gearboxes are an extremely efficient means of transmitting power. Depending

on the operating conditions, design, and manufacturing techniques, the loss per
mesh of spur, helical, or bevel gears will vary between approximately Vi to 3%.
Analytical means are available to calculate gearbox efficiency, but in order to be
confident that an accurate measure of a unit’s efficiency is known, the value
must be arrived at by testing.
Obviously, the efficiency of the gearbox affects the total system efficiency
and the rating of the prime mover required. The power lost in the gearbox, how­
ever, also has a large impact on system design since it must be dissipated as heat.
Many low-speed gearboxes have a self-contained splash lubrication system
and the power rating of the unit may be limited by the oil temperature rise
rather than a mechanical limit. This is the reason gearbox catalogs list a thermal
rating in addition to a mechanical rating. When the heat generated in a unit is
greater than can be dissipated through the casing an external lubrication system
including a cooler is required. The heat generation or efficiency of the gearbox
will determine the size of cooler required.
The heat generation can be calculated as follows:
Q = MCp AT
where

Q = heat generated, Btu/min

Cp = specific heat, Btu/lb-°F
 Table 11.3 Common Sources of Airborne and Structure-borne Sounds Generated in Gear Drive Systems21
Instruments that provide the operator with not only the amplitude of the vibration or noise, but also the predominant frequencies

356
can be a tremendous aid in determining sources. These causes normally present themselves as follows:
1. Balance: Residual unbalance presents itself at a frequency 7. Resonances: These also display themselves as frequencies
equal to once per shaft revolution and it will increase in ampli­ at which system members like to vibrate. The distinction from
tude as speed is increased. critical speeds is that resonances occur in other than rotating
2. A lignm ent: Misalignment will present itself at once or some­ members, and affect alignment. Resonances occur at fixed
times twice and three times per shaft revolution. However, the frequencies and change in amplitudes with load and temper­
amplitude will remain fairly constant with speed changes. ature.
3. Friction: This is difficult to pinpoint by vibration and 8. Tooth mesh (i.e., to o th co ntact): This will show up at
noise frequency. Amplitude may be very high when contin­ tooth mesh frequency (i.e., rotating speed times number of
uous sliding occurs. It may also be random, high-amplitude, teeth) and multiples of this mesh frequency.
shock-type pulses, as in hydrodynamic bearing rubbing. It 9. Bearing instability: Bad antifriction bearings will cause
may be irregular and often violent. high-frequency vibration at several times rotational speed, also
4. Looseness: This may cause unbalance, misalignment and friction vibration will occur. Hydrodynamic bearings, lightly
friction rubbing at moderate and high speeds. At low speeds, loaded, will tend to whirl at 0.43 to 0.47 times the rotational
it may display itself as an irregular rattle. Often it shows up at speed. This so-called “half-frequency whirl” will “on-set”
twice shaft rotational speed. violently with speed or temperature changes and may continue
until the rotor is completely stopped.
5. D is t o r ti o n : This is often an indirect cause of vibration and
noise, which also leads to unbalance, misalignment, or friction. 10. S ystem pulses: These may occur in many types of sys­
It will tend to change in amplitude with load or operating tems, such as the vane-pass frequency of a pump or compres­
temperatures, when speed is held constant. sor (rotational speed times the number of vanes), and the
6. Critical speeds: These occur through any given speed range beating of reciprocating engines which cause frequencies at one-
half and one-quarter rotational speed at various amplitudes.
and are points at which a rotating system likes to vibrate torsion-
ally or laterally at a particular frequency. Rotors characteristic­ 11. Windage: Couplings and other rotating parts generally
ally show violent increase in amplitude at particular critical create broadband noise, but can be at a bolt pass frequency or
speeds but are fairly stable above and below these speeds. A crit­ fan blade pass frequency.
ical speed may change frequency with load and temperatures.
aAll of these types of vibrations and noise frequencies can be generated in a gear drive. Major frequencies can interact and cause frequency modula­
tion and phase shifts. Any combination, sum difference, and multiple (harmonics) of prime frequencies can occur if the forcing magnitude and
system freedoms are such that they will cause and allow the generated vibration to become predominant. Generally, only the prime frequencies will
present themselves as problem modes. However, sometimes very elusive frequencies appear, such as periodic cutting machine error appearing on one
of the gears.
Source: Ref. 3.
Gear Unit Operation 357

M = oil flow, lb/min

AT = oil temperature rise across the gearbox, °F
Note:
Btu/min
—--------- = horsepower
42.44
Cp of oil = 0.5 Btu/lb-°F
1 gpm = 7.5 lb/min of oil
By accurately measuring the temperature of the oil entering andleaving the
gearbox and the oil flow, the heat rejected to the oil can be calculatedas shown
above.
For example, a gearbox with oil in temperature of 130°F, oil-out temper­
ature of 160°F and oil flow of 20 gpm will reject 848 Btu/min (20 hp) to the
oil. If the gearbox is transmitting 1000 hp, the efficiency can be calculated as
follows:
Pt - Pi 1000 - 20
E = J i = -------------- = 98%
Pt 1000
where
Pt = total hp transmitted
Pj = power loss
It must be realized that in addition to the heat rejected to the oil, heat is dissi­
pated to the atmosphere by radiation and convection through the casing. To
arrive at an estimate of the heat dissipated through the casing, the following heat
transfer equation may be used:
H = CA(Tc - Ta)
where
H = heat dissipated by the gear casing, Btu/hr
C = combined coefficient of radiation and convection, Btu/hr/ft/°F
Tc = casing temperature, °F
Ta = ambient temperature, °F
The value of C depends on the material, roughness, and color of the housing
and the velocity of air around the housing and can vary from approximately
0.5 to 3.0. To eliminate the variable of casing radiation and convection, the gear
housing can be insulated for an efficiency test. In this way all the heat generated
in the unit will be carried away by the oil.
358 Gear Drive Systems

The heat balance is the most widely used method of measuring gear unit
efficiency. A more direct way is to measure input and output torque, but this
method requires exact and expensive instrumentation. In the case of a regener­
ative or back-to-back test, only input torque need be measured and this will
monitor the torque required to drive the two locked-up gear trains. Since the
prime mover supplies only enough power to make up the losses in the system,
the torque lost in each gearbox is one-half the input torque.
Power losses in the gearbox can be divided in two categories, friction losses
in the gear mesh and at the bearing interfaces and windage or churning losses.
In high-speed boxes, windage and churning losses can be considerable amounting
to up to half the total power loss. An easy way to determine these losses is to
operate the gearbox at full speed, no load, and then perform a heat balance.
A high-speed box is considered to be one with pitch line velocity ranging
from 5000 to 20,000 fpm. Above 20,000 fpm windage and churning can be a
limiting factor to the success of the design. For instance, some wide-face-width
helical gearboxes operating at pitch line velocities above 20,000 fpm have
experienced severe heat generation problems due to pumping of oil and air along
the face width. In high-speed double helical gears the hands of helix should
always be selected so as to pump oil away from the apex.
When it is determined that a gearbox has excessive windage and churning
losses, the situation can be relieved by strategically placing baffles or screens in
the unit. In low-speed units using splash lubrication it is beneficial to shroud the
gears dipping into the sump oil. With high-speed units the sump oil level should
be below the gears and the sump should be separated from the gear meshes by a
baffle plate or screen. Minimizing windage and churning losses is more an art
than a science and can require trial-and-error solutions verified only by testing.
Lubrication system parameters can have a profound effect on gearbox
efficiency. For instance, efficiency will increase with reduced oil flow; obvi­
ously, this is due to reduced churning. The limit to this technique is the increase
in temperature rise across the box and in the individual components as oil flow
is reduced. Increasing oil in temperature will increase efficiency since the oil
viscosity will decrease and therefore reduce churning. Changing to a lower-
viscosity oil will have the same effect. Proper scavenging of the unit is extremely
important in minimizing churning; therefore, great care must be taken in the
location and size of drain holes. Also, back pressure must be minimized.

INITIAL FIELD STARTUP

Prior to starting the equipment, the following preliminary checks should be

performed:
Gear Unit Operation 359

1. Check oil level and ensure that the proper oil is being used.
2. Tighten all pipe connections.
3. Check all electrical connections.
4. Tighten all mounting and gearbox bolts with proper torque.
5. Check mounting of all gauges, switches, and so on.
6. Check all couplings for proper installation and alignment.
7. Check inspection cover installations.
The following instructions pertain to the initial startup:
1. The unit should be preoiled to ensure lubrication of the journal bearings at
startup.
2. The gearbox should be started slowly under as light a load as possible.
Observe that the rotation is in the proper direction. Check the system oil
pressure.
3. After starting, when the oil has been circulated, the unit should be stopped
and sufficient oil added to bring the sight gage oil level up to the specified
amount.
4. As the unit is brought up to operating speed, it should be continuously
monitored for excessive noise, vibration, or temperature. If any of these
occur, shut down immediately, determine the cause, and take corrective
action. Also check for oil leaks as the unit is initially operated.
5. If possible, operate at half load for the first 10 hr to allow final breaking in
of the gear tooth surfaces.
6. After the initial 50 hr of operation the oil in a new unit should be drained
and the case flushed with SAE 10 straight mineral flushing oil containing no
additives. Drain the flushing oil and refill with the recommended lubricant to
proper level.
7. After the initial 50 hr of operation check all coupling alignments and
retorque all bolts. Check all piping connections and tighten if necessary.
If starts are made in a cold environment, consideration should be given to pre­
heating the lubricant. Load should not be applied until the lubricant has attained
operating temperature.

CONDITION MONITORING

On an operational gearbox the question always arises as to how to evaluate the

condition of the unit and when to disassemble the equipment and inspect or
replace internal components. Figure 11.16 presents a hazard function or what is
traditionally called the “bathtub curve,” which describes the failure rate of a
unit at any particular point in its operating history. Initially, there is a high
“infant mortality” period, which reflects failure due to assembly or
360 Gear Drive Systems

TIME ON COMPONENT

Figure 11.16 Hazard function.

manufacturing errors. For instance, an oil jet might have been clogged or a
bearing preloaded at assembly. These early failures can often be screened out
during an acceptance test program. Following the “infant mortality” phase,
there is a period of constant failure rate where failure modes related to time
appear. Examples of these are bearing or gear tooth fatigue, spline wear, and so
on. At first glance it would appear reasonable that the unit be removed from
service at the point in time where the failure rate begins to increase and the
components inspected and reworked or replaced as necessary. This is not
necessarily so, since removal and disassembly of the unit begins another operat­
ing cycle and the gearbox again would be subject to the high “infant mortality”
failure rate. As a practical matter it is difficult to establish the point where the
failure rate increases sufficiently to warrant overhaul since this requires an
extensive data base. Gear units are subject to many different modes and within
any given failure mode there will be significant scatter.
The most efficient method of determining when a gear unit requires
service is to base repairs on the condition of the gearbox rather than overhaul
it at an arbitrary time period. The basic idea is not to disturb equipment that is
operating properly but inspect only items that exhibit potential failure
symptoms. This type of predictive maintenance is termed “on condition” and
requires instrumentation of a gear unit and the proper interpretation of the data
provided. The goals of this technique are threefold:
1. Detect gear units which are operating abnormally.
2. Diagnose which internal component is deteriorating.
3. Predict how long the unit can function in this condition before corrective
action must be taken.
Gear Unit Operation 361

Condition monitoring systems are expensive and require discipline to set up and
use effectively. It is difficult to define the normal operating parameters of a
system and this initial step is vital in an “on condition” program. Sometimes the
normal operating parameters are termed the “signature” of a component. Once
this is established it is relatively easy to identify abnormal operation. When an
“on condition” program is achieved, the benefits are as follows:
1. Reduction in downtime since potentially catastrophic failures are detected
and corrected. This also leads to a reduction in repair costs.
2. Enables efficient scheduling of equipment shutdown since warning of a
failure is detected well in advance of the event.
3. Shutdown periods are shorter because unnecessary work is avoided and the
work required is known in advance so that preparations can be made.
In the following pages the various methods used to monitor gearbox condition
will be discussed. These include:
Vibration
Noise
Shaft position
Oil temperature
Oil pressure
Oil analysis
Chip collection

Oil Analysis

Gearbox condition can be monitored by the analysis of oil samples. One widely
used technique is SOAP (Spectrographic Oil Analysis Program). An oil sample is
taken periodically from the gearbox sump and sent to an analytical laboratory,
where it is burned and the light waves passed through a spectrometer. The
spectrum of light waves given off by the sample yields information as to the
types of wear metals suspended in the oil and their quantity.
In principle, if it is possible to identify and measure the wear metals
present in the lubricant, one can identify which internal gearbox component is
deteriorating. Also, by monitoring the rate of change of wear metal from sample
to sample, conclusions can be drawn concerning the gearbox condition and when
maintenance will be required. Figure 11.17 illustrates the information presented
in a typical SOAP laboratory analysis.
To use SOAP properly, a gearbox history base must be established to
determine what type and quantity of wear metals the unit normally generates.
Also, a listing of materials used in the gearbox must be available. With this infor­
mation SOAP can be used to predict failures. For instance, an increase in silver
might mean that a silver-plated antifriction bearing retainer is wearing. Tin and
 Z9£
5 SAMPLED
O
V
H
8 m R EC E IV E D m
*8. OIL
m
Figure 11.17

few W x
o
v «* c
UNIT 3D
U* fea CO
I I
OIL
ADDED

<* *& VIS. AT 10 0 ° F

O O T O T A L ACID
bi*» b NO.
Typical oil analysis form.

A A
"ESTS

o WATER
OPERT\

«*
iY S IC A L

*
© o IRON
o 0 LEAD
>- w
o COPPER 2
b m
o o H
►» CHROMIU M >
r;c o
CO "0
© o ALUMINUM • • in
> o
© NICKEL jo zj
o H 5
CO o
o b SIL V ER _ o
-o
5m
o b TIN * 2
o
r © SILIC ON
(-
nr- >
r o BORON o >
Z Z
©p PHOSPHORUS
*< ^
v# SODIUM
b m m
h ZINC Q
o 0 X
H
.5 b CALCIUM
*> BARIUM
o o
0 MAGNESIUM
o
TITANIUM
O b
Gear Unit Operation 363

lead traces could indicate a journal bearing problem. A large percentage of iron
might point to gear or bearing distress. Of course, practically all gearbox compo­
nents contain some iron, but if other elements are also indicated, the cause of
distress might be narrowed down to a specific component.
SOAP oil samples should always be taken from the same area of the unit.
Oil sampling periods are in the order of every 10 to 100 hr of operating time
and each oil sample should be placed in a clean vial marked with an identifying
number, the date, and the number of operating hours of the unit. If readings
indicate a problem, the sampling interval should be reduced. There is a time
delay in getting results from the laboratory and a typical turnaround time might
be 3 days.
SOAP readings can be influenced by external factors. For instance, oil
changes or oil additions must be accurately recorded since they will reduce the
percentage of wear metals. Also, the degree of filtration will affect the wear
metals suspended in the fluid. Very fine filtration such as 3 pm will effectively
remove all wear metals and make SOAP meaningless. It should also be recog­
nized that large particles will settle out and not be detected by SOAP analysis.
The effective detection range of the SOAP procedure is with particles in the size
range 1 to 10 pm.
Increases in the wear metal percentage between samplings are more signifi­
cant than the total wear metal content since an increase signifies a trend of
abnormal operation. Therefore, a rapid increase in wear metal should be investi­
gated even if the absolute percentage does not exceed the acceptance criterion.
When performing a SOAP analysis other oil characteristics are commonly
monitored, such as acid number, viscosity, and water content. The acid number
is relatively easy to check and will increase due to the presence of wear metals
or due to overheating. Occasionally, one can detect acid oil by the odor alone,
which is distinctive and unpleasant. Changes in oil viscosity indicate lubricant
deterioration and can be monitored. Monitoring of acid number and/or viscosity
can be done more quickly and economically than SOAP but does not yield as
much information.

Vibration

Gear boxes are mass elastic systems and therefore will vibrate when excited by
internal or external influences. Vibration of gear units can be measured in several
different ways. The motion of the shafts, both radial and axial, can be observed
with noncontacting sensors. Casing motions can be measured using velocity or
acceleration pickups. Generally, the following parameters are monitored:
1. Peak-to-peak mils displacement
2. Peak inches per second velocity
364 Gear Drive Systems
+

Figure 11.18 Simple harmonic motion.

3. Peak g’s acceleration

4. Frequency of vibration
The relationship of vibratory displacement, velocity, and acceleration is as
follows:
A vibration of simple harmonic motion which has a pure sinusoidal wave­
form has a displacement d (see Figure 11.18):
d = d0 sin cot
where
d0 = % peak-to-peak displacement reading, in.
co = frequency, rad/sec
= (27r)rpm/60
t = time, sec

Differentiating for the velocity v:

v = co d0 cos cot
and differentiating once more for the acceleration a:
a = -co2 d0 sin cot
To calculate the G loading, let sin cot = 1, and divide a by the acceleration of
gravity:
Gear Unit Operation 365

where g is 386 in./sec2.

The major contributors to gearbox vibration are:
1. Rotating shaft unbalance
2. Shaft assembly problems such as loose connections, bent shafts, and
misalignment
3. Gear tooth inaccuracies
4. Component wear
AGMA Standard 426.01 [4] presents acceptable vibration levels, as shown in
Figure 11.19. If the displacement amplitude is not obtainable at discrete fre­
quencies, the standard allows either of the following:
1. A nominal unfiltered velocity level of 0.3 in./sec but not exceeding a maxi­
mum 2 mil displacement. {Note: a mil is 0.001 in.)
2. An unfiltered displacement level determined from Figure 11.19 using the
shaft rotation speeds as discrete frequencies (i.e., 120 rpm = 20 Hz).
A gearbox may be acceptable according to the criteria of Figure 11.19 in accep­
tance testing at the manufacturer’s facility, yet exhibit higher vibration levels in
field service. The vibration levels may be adversely affected by factors not under
control of the gear manufacturer, such as:
1. Inadequate foundation
2. Excessive shaft misalignment
3. Coupling components not tested with the gear unit
4. Resonance of base or other supporting structure
5. System torsional vibration
6. Motor magnetic center wobble
7. Unbalance or other forced vibration from other components in the system
Shaft vibration levels as shown in Figure 11.19 are measured by the use of non­
contacting, eddy current proximity devices. A probe is positioned approximately
0.050 in. from a shaft and the distance (gap) between the probe tip and the shaft
is translated to a voltage.
A proximity probe is basically a small coil of wire embedded in a ceramic
tip on the end of a holder. This coil is supplied with a high-frequency voltage
which produces a magnetic field radiating from the probe tip. Any conducting
surface that lies within this magnetic field absorbs some of its power, and the
closer to the tip the more power absorbed. The RF signal to the coil is supplied
by a proximitor which has the secondary function of relaying the resultant field
strength back to the readout instruments by reconverting it to a voltage.
Proximitors and probes are calibrated to give a linear relationship between
readout voltage and distance of observed material. The range of this linear rela­
tionship depends on the type of probe and supply voltage, but is on the order of
 366 Gear Drive Systems
PEAK- PEAK
AMPLITUDE-MILS

FREQUENCY ( H i)

Figure 11.19 Acceptable shaft vibration levels. (From Ref. 4.)

80 mils. The readout obtained from a radially mounted probe is the electrically
measured difference between the minimum and maximum distances of the probe
from the shaft material (peak-to-peak amplitude).
The American Petroleum Institute has issued a standard [5] which defines
the use of proximity probes. Figures 11.20 to 11.22 illustrate radial and axial
probe arrangements and their application in a gearbox.
Note that in Figure 11.22 two radial probes 90° apart are mounted at the
bearing positions. It is possible to have totally different vibration in two
 Gear Unit Operation 367

g f /I MU

CORD CLAMP CASING OR BARREL

CLAMPED TO BEARING
OR BEARING HOUSING

BEARING OR CASING

USE LOCTITE
DROGUE
NYLOC ON
FOR
BLIND GUIDE THREADS

OPTIONAL: RESTRAINT: STANDARD OPTIONAL

REVERSE MOUNTED FOR LONG LENGTHS LO CTITE JAM NUT T IE D
PROBE W ITH R E S TR A IN T
ON PROBE HOLDERS
ON THREADS
(NON-CLAMPING)

Figure 11.20 Typical radial probe arrangement. (From Ref. 5.)

368 Gear Drive Systems

Figure 11.21 Typical axial probe installation. (From Ref. 5.)

perpendicular directions at one particular bearing. This is generally true of any

vibration measurement, whether it be at the bearing or at the casing and readings
in two planes should always be taken. Gearbox bearing probes are mounted 45°
from the vertical center since quite often there is a split line at the horizontal.
On Figure 11.22 a phase angle probe is shown on each shaft. This is a
transducer that observes a once-per-turn event such as a key way. Its function is
to provide a reference mark and timer for speed, phase angle, frequency
measurements, and all data acquistiion. Figure 11.23 shows a phase angle probe
installation. Each time the keyway passes the probe a voltage pulse results. This
pulse provides a physical reference on the shaft which can be used to measure
the high spot of the shaft.
 Gear Unit Operation 369

ITEM DESCRIPTION

( J l j .75 CONDUIT ^ f g)|. 1 Primary Axial Position Probe.

Vendor and Model Number.
Target Area.
2 Secondary Axial Position Probe.
Vendor and Model Number.
Target Area.
3Y Input Shaft End Radial
Vibration Probe. 45° Off TDC.
Vendor and Model Number.

4X Input Shaft End Radial

Vibration Probe. 45° Off TDC.
Vendor and Model Number.
5Y Output Shaft End Radial
Vibration Probe. 45° Off TDC.
Vendor and Model Number.

6X Output Shaft End Radial

Vibration Probe. 45° Off TDC.

r I
M =J»
n 0
1&2
R
Vendor and Model Number.

Phase Angle Probes (2).

Vendor and Model Number.

Radial Bearing - Description

T Thrust Bearing - Description

® — - X X JH
JB
Fitting - Description

Junction Box - Description

CONDUIT
(JH) VIBRATION AND AXIAL POSITION MONITOR

ccw AXIAL RADIAL RADIAL

ROTATION POSITION VIBRATION VIBRATION
1
VIEWED
HERE Input Output
Shaft Shaft

n ...._kl * ..... 2L Y X_
#1 #2 #3 #4 #5 #6

Figure 11.22 Typical system arrangement for double helical gear. (From
Ref. 5.)
370 Gear Drive Systems

typical output

PROBE AND HOUSING ASSEMBLY WEATHERPROOF

PROBE LEAD -HOUSING

PRO> IMITOR
EXTENSION SEALTITE
3-W IRE SIG NAL CABLE
CABLE r OR
/ CONDUIT

TO MONITOR
-M A C H IN E CASE
PROBE

PROBE GAP
KEYWAY

SHAFT

Figure 11.23 Phase angle probe. (Courtesy of Bently Nevada, Minden, Nevada.)
Gear Unit Operation 371

To illustrate the type of data acquired by proximity probes, refer to

Figure 11.24. The oscilloscope photograph displays unfiltered time-domain
traces from both horizontal and vertical probes, as well as the same signals
filtered at rotational frequency. The inclusion of a phase angle mark, K0 ,
identifies one complete revolution of the shaft. In the middle of the picture
the shaft orbit can be observed as an unfiltered motion, and filtered exactly at
running speed. In addition, a spectrum analysis of each signal is presented, and
the various components identified.
There are two types of oscilloscope readings presented in the center of
Figure 11.24. One is the time-base mode, where the sinusoidal-type waveform
representing the shaft motion is displayed. This shows the position of the shaft
relative to the input transducer versus time horizontally across the cathode ray
tube of the oscilloscope. The other display is the orbit presentation where the
output from two separate proximity probes at 90° to one another are shown in
the X-Y mode of the oscilloscope. In this mode the centerline motion of the
shaft is displayed. If the probes are mounted at the bearing, the orbit is a presen­
tation of the motion of the shaft centerline with relationship to the bearing.
When measuring shaft vibration with proximity probes it is possible to get
erroneous readings caused by physical and mechanical deformities in the shaft
material. These erroneous readings are sometimes referred to as glitches.
Obviously, there will always be some mechanical runout in the shaft. This
may be measured by a dial indicator mounted in the probe area of the shaft,
which itself is supported at the bearing journals in V blocks or rollers. The shaft
runout at the probe locations may be subtracted from the vibration readings to
arrive at actual vibration levels, provided that the runout is shown to be in phase
with the vibration.
In addition to runout, other mechanical problems that will cause glitches
are bowed shafts and surface imperfections such as scratches, dents, and burrs.
To correct these problems the shaft surface may have to be remachined or
reground. If the surface irregularities are very minor, a redressing of the surface
with an oil-wetted fine emery cloth may eliminate the surface imperfections.
In addition to mechanical runout there is the possibility of electrically
induced runout due to residual magnetism which can be corrected by having the
shaft degaussed. Other potential causes of glitch could be metallurgical segrega­
tion or residual stress concentrations.
Before assmbling a shaft that has probe areas a bench test should be con­
ducted to ensure that glitch has been held to minimum (in the order of 0.00025
in.) so as not to influence vibration readings excessively. After functional inspec­
tion it is advisable to coat the probe areas with an epoxy resin which can remain
in position for the life of the machine. This coating will not affect probe
readings but will protect the probe area from corrosion and minor mechanical
damage. The shaft position probes provide a vibration measurement of the
to peak)]
[mil (peak
DISPLACEMENT
RADIAL
ROTOR

Figure 11.2 4 Shaft displacement data. (Courtesy of Bently Nevada, Minden,

Nevada).
Gear Unit Operation 373

INTERCONNECTING
CABLE

VELOCITY-TO-DISPLACEMENT VELOCITY SEISMOPROBE

CONVERTER

Figure 11.2 5 Velocity transducer system. (Courtesy of Bently Nevada, Minden,

Nev.)

relative motion between the shaft and the mounting of the proximity probe. The
probe is usually mounted rigidly to a bearing cap or the gear casing. In order to
establish motion of the casing itself, velocity pickups are attached to the casing.
Figure 11.25 illustrates a velocity transducer system. The purpose is to
measure gearbox casing or structural vibration velocity and convert the vibration
velocity into an electrical signal that represents the displacement of the casing.
One type of velocity transducer works on the principle that as a magnet
moves with respect to a relatively stationary coil, a current is induced in the coil.
The magnet is rigidly attached to the pickup case and therefore vibrates along
with the casing. The coil is suspended by sensitive springs inside the pickup case.
Figure 11.26 depicts the casing velocity characteristics emitted by a three-
stage speed-increasing gearbox. The time-domain signal inset in the figure is quite
complicated due to the transducer responding to a large number of excitations.
A spectrum analysis of the signal reveals Fourier components that can be related
back to the operational characteristics of the machine. Peaks can be seen at the
rotational speeds of each gear plus the second harmonics of each fundamental
frequency. As a rule of thumb, peak vibration velocity readings over approxi­
mately 0.5 in ./sec signify extremely rough operation and warrant shutdown of
the machinery. Readings below 0.1 in./sec indicate smooth, well-balanced, and
well-aligned equipment. A reading of over 0.3 in ./sec might be a good level at
which to consider taking corrective action.
Casing vibration should be measured in the vertical, horizontal, and axial
directions. The measurements should be taken on a rigid section on the housing.
It should be noted that a measurement taken on a rigid casing can identify such
things as structural or piping resonance, loose or cracked foundations, or
external vibration input sources but may not transfer vibration amplitudes due
 374 Gear Drive Systems
(zero to peak)]
[in./sec

0.150
VELOCITY

0.125
CASING

0.100

0*075

0*050

0*025

FREQ UENCY (cycles/Min x K T *)

Figure 1 1.26 Gear box casing velocity characteristics. From Ref. 6; courtesy
of Bently Nevada, Minden, Nev.)
Gear Unit Operation 375

Table 11.4 Vibration Conversion Factors

Multiply the following by the numerical values:
To obtain Average Rms Peak . Peak to peak
Average 1.0 0.900 0.636 0.318
Rms 1.111 1.0 0.707 0.354
Peak 1.571 1.414 1.0 0.500
Peak to peak 3.142 2.828 2.0 1.0

to shaft motions. The casing may be too stiff to move as a result of shaft
motion; therefore, on critical applications shaft position sensors should be incor­
porated in addition to casing pickups. When a casing transducer is located in the
same plane as a shaft proximity probe, a vector summation of the two outputs
can give the absolute shaft motion in addition to the shaft motion relative to the
casing.
On occasion, rather than defining vibration in peak values, rms or average
values are required. Table 11.4 presents the conversion factors. For high-
frequency excitations such as gear meshing frequencies, accelerometers are used
as transducers. Figure 11.27 shows an acceleration transducer system. The
accelerometer portion of the acceleration transducer system is a “contacting”
transducer that is physically attached to the vibrating machine part. The
accelerometer uses a piezoelectric crystal situated between the accelerometer

Interface Module Accelerometer

Figure 11.27 Acceleration transducer system. (Courtesy of Bently Nevada,

Minden, Nev.)
 376 Gear Drive Systems
to peak)]
[Gs(zero
ACCELERATION
CASING

Figure 11.28 Gear box casing acceleration characteristics. (From Ref. 6;

courtesy of Bently Nevada, Minden, Nev.)
Gear Unit Operation 377

base and an inertial reference mass. When the crystal is strained (compression or
tension force), a displaced electric charge is accumulated on the opposing major
surfaces of the crystal. The crystal element performs a dual function. It acts as
a precision spring to oppose the compression or tension force and it supplies an
electric signal proportional to the applied force.
Frequencies up to 30 kHz can be handled by accelerometers. Figure 11.28
illustrates gearbox casing acceleration characteristics. The top spectrum plot
depicts a gearbox in good mechanical condition with reasonably low acceleration
levels and a normal mixture of components. A similar measurement made on a
unit that had cracked pinion teeth is presented on the bottom plot. High G
loadings are exhibited at the gear mesh frequency, the seventh harmonic of
pinion rotational speed, and the pinion running speed, XHS.

Figure 11.29 Magnetic chip detector. (Courtesy of Technical Development

Company, Glenolden, Pa.)
378 Gear Drive Systems

i /111
I J
Figure 11.30 Chip collector with wear debris. (Courtesy of Technical Develop­
ment Company, Glenolden, Pa.)

Chip Collection

A typical magnetic chip collector is shown in Figure 11.29. Figure 11.30 is a

photograph of a chip collector with typical wear debris. The chip collector is
installed in the oil sump or a scavenge return line. Figure 11.31 shows typical
installations. A self-sealing valve allows withdrawal of the magnetic probe and
visual inspection of the collected debris with loss of only a few drops of oil.
A system can be initiated where the chip collector is periodically in­
spected, every 25 to 50 hr, and records kept characterizing the particles
collected. Once a baseline is defined, any change in the debris collected during a
Gear Unit Operation 379

Figure 11.31 Chip collector installations. (Courtesy of Technical Development

Company, Glenolden, Pa.)

given interval, such as rate of collection or size of particle, identifies a potential

problem. The particles can be analyzed by electron spectroscopy to determine
what elements are present and yield clues as to which components are deterior­
ating [7].
Because the particles trapped are conductors the magnetic probe can be
designed such that a gap is bridged between two electrodes when sufficient
particles land on the sensor. When this occurs a warning light can be activated.
Figure 11.32 shows two such electric chip detectors with radial and axial gaps.
380 Gear Drive Systems

Figure 11.32 Electric chip detector. (Courtesy of Technical Development

Company, Glenolden, Pa.)

This type of detector is widely used in aircraft and turbine engine transmissions.
Some development of the proper gap size is required since too small a gap will
lead to many “nuisance” indications and too large a gap will be insensitive to
failure indications. Gaps used are on the order of 0.050 to 0.150 in. With chip
collectors or detectors care must be taken to locate them in an area where they
will be exposed to as many particles as possible. Also, the area should be rela­
tively still, so the particles can settle.
In forced-flow lubrication systems where particles tend to be dragged
along with the oil system, full-flow debris detectors are sometimes used. These
configurations have screens through which the total scavenge oil flows. The
screens retain the debris. A chip detector can be incorporated for indication pur­
poses. Figure 11.33 shows a full-flow debris detector. An advantage of this type
of monitor is that nonmagnetic as well as magnetic debris will be trapped.
Another type of full-flow monitoring device is the indicating screen
(Figure 11.34). It is woven from wire strands and when a conductive particle
bridges the gap between wires, a warning light can be activated. This device can
be used as a pump inlet screen. The indicating screen is sensitive to magnetic and
nonmagnetic debris but is more difficult to remove and inspect than a chip
detector.
Gear Unit Operation 381

Figure 11.33 Full-flow debris collector. (Courtesy of Technical Development

Company, Glenolden, Pa.)

Oil Pressure and Temperature

Oil pressure and temperature measurements are relatively easy to accomplish and
yield significant information concerning the operating condition of a gear unit.
The oil pressure should be monitored at the entrance to the unit down­
stream of any pressure regulation device. Low oil pressure can indicate internal
leakage such as a cut static seal or possibly low flow due to pump distress. This
can lead to oil starvation of components; therefore, low oil pressure must be

Figure 11.34 Indicating screen debris monitor. (Courtesy of Technical Develop­

ment Company, Glenolden, Pa.)
382 Gear Drive Systems

investigated and corrected. High oil in pressure can be an indication of down-

stream blockage such as a clogged oil jet and therefore must also be investigated
and corrected. A fluctuating oil pressure can be an indication of pump cavitation
due to poor inlet conditions such as excessive pressure drop or air leakage in the
pump inlet line. When setting oil pressure limits it must be remembered that the
pressure will be affected by variables such as oil temperature and viscosity and
jet size variation downstream of the pressure measurement. Pressure in the gear­
box cavity will also directly affect the feed pressure.
Oil temperatures should be monitored at the gearbox inlet and outlet at
a minimum. The change in temperature across the unit is a good indication of
gearbox condition. Scavenge oil temperature of individual bearings can be
monitored. On journal bearings, temperature sensors can be embedded under
the surface of the babbitt at a depth of 0.030 to 0.060 in. This is a very positive
way of monitoring bearing condition. More than one sensor should be installed
in case of sensor failure.
Two types of temperature sensors are available: the resistance temperature
detector (RTD) and the thermocouple. The RTD is basically a precision resistor
where the resistance changes linearly with temperature. RTDs are superior to
thermocouples in terms of accuracy, stability, and interchangeability.

REFERENCES

1. AGMA Sound Manual 299.01, Sec. I, Fundamentals of Sound as Related to

Gears, American Gear Manufacturers Association, Arlington, Va., May 1978.
2. AGMA Standard 295.04, Specification for Measurement of Sound on High
Speed Helical Gear Units, American Gear Manufacturers Association, Arling­
ton, Va., April 1977.
3. AGMA Sound Manual 299.01, Sec. II, Sources, Specifications and Levels of
Gear Sound, American Gear Manufacturers Association, Arlington, Va.,
March 1980.
4. AGMA Standard 426.01, Specification for Measurement of Lateral Vibration
on High Speed Helical and Herringbone Gear Units, American Gear Manu­
facturers Association, Arlington, Va., April 1972.
5. API Standard 670, Noncontacting Vibration and Axial Position Monitoring
System, American Petroleum Institute, Washington, D.C., June 1976.
6. Bently Nevada Application Notes, Bently Nevada Corp., Minden, Nev.
various dates.
7. Tauber, T., A Design Guide to Effective Debris Monitoring in Gas Turbine
Engines and Helicopter Transmissions, Technical Development Company,
Glenolden, Pa.
12
MAINTENANCE AND FAILURE ANALYSIS:
SCHEDULED MAINTENANCE ACTIONS

The objective of a maintenance program is to ensure satisfactory gearbox per­

formance at all times and to maintain the transmission in a state of readiness if
it is not in operation. A program should be planned which includes regular
maintenance actions and periodic monitoring of operating characteristics to
determine whether any problems are developing.
Every gearbox should come with a set of maintenance instructions devel­
oped by the manufacturer. These instructions should include the following
information:
General description of the equipment
Specifications as to speed and power ratings, lubrication, dimensions, and so on
Installation information
Lubrication instructions
Maintenance requirements
Assembly and disassembly instructions
Drawing and parts list
The following section presents some general comments on maintenance
of gearboxes. A specific unit may require special maintenance procedures and
the manufacturer’s instructions should be adhered to.
There are several maintenance actions that should be affected during the
initial operation of a gearbox. After approximately 50 hr of running time, the
following should be accomplished:
Check coupling alignment and correct if necessary.
Check bolt torques and retighten if necessary.

383
384 Gear Drive Systems

Table 12.1 Regular Gearbox Maintenance Actions

Frequency Maintenance item Corrective action
Daily Check oil temperature If there is a drastic
and pressure at oper­ change from previous
ating conditions. readings, stop unit
and determine cause.
Check for noise, vibration,
and oil leaks.
Check sump oil level. Add oil if necessary.

Weekly Check oil filter. Change filter element

if necessary.

Monthly Check lubricating oil for Drain and refill lube

contamination. system if necessary.
Change oil filter.
Check all gages, controls,
and alarm systems.
Clean breather element.

Every 2500a Change lubricating

hours or system oil.
6 months
aIf operating conditions are unusually severe, such as high-temperature or high-moisture
atmospheres, oil change requirements might be more frequent. Changes can be based on
inspection of the oil for viscosity or acid number in such cases.
Source : Sawyers Turbomachinery Maintenance Handbook, 1st ed., Vol. Ill, Turbo­
machinery International Publications, Norwalk, Conn., 1980.

Check piping connections and retighten if necessary.

Change oil and clean sump.
The oil, after 50 hr, need not be discarded but can be drained, filtered through
an element with a micrometer rating no greater than the gearbox filter, and
reused. Particles may be found in the oil and the sump due to normal wearing in
processes. At this point the sump or reservoir should be thoroughly cleaned.
After draining the original oil it is recommended that the gearbox and lubrica­
tion system be flushed out with a flushing oil. If possible, bring the unit up to
operating speed at light load after filling with flushing oil. Shut down immedi­
ately after achieving full speed. The drain the flushing oil and refill with the
recommended lubricant to the proper level.
After the 50 hr maintenance a regular program should be followed as out­
lined in Table 12.1. Logs should be kept of instrument readings and maintenance
actions to keep a running account of gearbox condition.
Maintenance and Failure Analysis 385

When performing maintenance operations every precaution must be taken

to prevent foreign matter from entering the gearbox. The introduction of mois­
ture, dirt, or fumes can lead to sludge formation and deterioration of the lubrica­
tion and cooling system.

STORAGE

Quite often the gearbox is delivered before the complete system is ready for
assembly and it must be stored for some period of time prior to operation. When
operation is delayed more than 1 month after shipment, special precautions
must be taken to prevent rusting of the components. If possible, the gearbox
should be completely filled with oil during storage. Where this is not practical,
all exposed metal parts, both inside and outside the unit, should be sprayed with
a heavy-duty rust preventative. The gearbox should be stored in a dry area
remaining at approximately constant temperature, preferably indoors. If stored
outdoors, the gearbox should be raised off the ground and completely enclosed
by a protective covering such as a tarpaulin. If possible, the unit should be
rotated at weekly intervals while in storage.

OVERHAUL AND SPARE PARTS

Generally, gearboxes do not have a specific time period after which the unit is
disassembled and overhauled. It is more common to observe deterioration of
components such as bearings and gears during operation and replace the par­
ticular component at a convenient time. Usually, the gearbox is delivered with
an operating and maintenance manual which describes how to disassemble and
assemble the unit. If the user is not completely familiar with the equipment, it
would be prudent to have a factory representative accomplish any major com­
ponent replacements. Spare gears or bearings for gearboxes are not necessarily
readily available from the manufacturers. Journal bearings, unlike antifriction
bearings, are not usually stocked by distributors. In many cases the bearings are
customized for the specific gearbox and therefore are even harder to replace.
Finished gears are rarely stocked by manufacturers, and lead times on gears or
bearings might be 20 weeks or more. When purchasing the drive the user should
request a recommended spares list and determine what the availability of these
parts will be. The user and manufacturer can then arrive at some agreement
over what spares will be available and where they will be stored.

TROUBLESHOOTING

The major causes of gearbox failure are improper lubrication and overload. Care
must be taken to check for proper oil level before operation. Excessive oil
386 Gear Drive Systems

Table 12.2 Gearbox Troubleshooting Chart

Recommended
Problem inspection Corrective action
Overheating 1. Oil cooler Check flow of coolant and oil flow.
operation Measure oil temperature into and out
of cooler. Check cooler internally for
buildup of deposits from coolant
water.

2. Is oil level too Check oil level indicator.

low or too high?

3. Bearing Make sure that bearings are not pinched

installation and properly adjusted.

4. Grade and con­ Check that oil is specified grade. Inspect

dition of oil oil to see if it is oxidized, dirty, or
with high sludge content.

5. Lubrication Check operation of oil pump. Make sure

system that suction side is not sucking air.
Measure flow. Check if oil passages are
free. Inspect oil line pressure regulator,
nozzles, and filters to be sure they are
free of obstruction.

6. Coupling float Check coupling alignment and adjust

and alignment end float.

Shaft failure 1. Type of Rigid couplings between rigidly sup­

coupling ported shafts can cause shaft failure.
Replace with coupling to provide
required flexibility and lateral float.

2. Coupling Realign equipment as required.

alignment

3. Excessive over­ Reduce overhung load. Use outboard

hung load bearing or replace with higher capacity
unit.

4. High transient Apply couplings capable of absorbing

loading shocks. Use couplings with shear
pins.
Maintenance and Failure Analysis 387

Table 12.2 (Continued)

Recommended
Problem inspection Corrective action
Shaft failure 5. Torsional or Adjust system mass elastic characteristics
(continued) lateral to control critical speed location.
vibrations Possibly, coupling geometry can be
modified.

6. Cracks due Note cause of fretting and correct. Press

to fretting fits between gear and shaft.
corrosions

Oil leakage 1. Exceed oil level Check oil level indicator.

2. Is breather open Check oil breather.

3. Are oil drains Check that all oil drain locations are free
open and clean.

4. Oil seals Check oil seals and replace if worn. Check

condition of shaft under seal and polish
if necessary.

5. Plugs at drains, Apply sealant and tighten fittings.

levels, and pipe
fittings

6. Housings and Tighten cap screws or bolts. If not effec­

caps tive, remove housing cover and caps.
Clean mating surfaces and apply new
sealing compound. Reassemble. Check
compression joints by tightening
fasteners firmly.
Source : Sawyers Turbomachinery Maintenance Handbook, 1st ed., Vol. Ill, Turbo­
machinery International Publications, Norwalk, Conn., 1980.

volume can be as detrimental as lack of lubrication and will result in churning

and overtemperature of components. Overload can be a result of vibration,
shock loads, or high torque at low speed. If there is a possibility that operating
loads will exceed rated gearbox loads, the manufacturer should be consulted.
Table 12.2 is a troubleshooting guide that gives some guidance in how to
identify and correct some of the common problem areas occurring during gear­
box operation.
388 Gear Drive Systems

Gear teeth should be inspected for nicks, burrs, and scratches, which may
be repaired by blending provided that they are minor and not on the working
surfaces of a tooth. The blend may be accomplished using a small file and an
India or carborundum stone. Crocus cloth should be used for the final polishing.
All repairs must be finished smoothly. Power tools are not permitted for blend
repairs.
In many gearboxes the teeth can be visually inspected by removing inspec­
tion covers bolted into the casing. When opening these inspection covers care
must be taken to ensure that no foreign material enters the gearbox. Gear teeth
should be examined under good lighting and be wiped clean of oil to prevent a
false diagnosis. The content pattern should cover approximately 80% of the
tooth. The following section discusses gear tooth failure modes and describes
various conditions that may be encountered in the field.

Gear Tooth Failure Modes

The major modes of gear tooth failure are breakage, wear, pitting, and scoring.
When these problems are encountered in the field it is important to accurately
define the condition and causes of failure to be able to determine corrective
actions for the particular units in difficulty and also to modify analytical and
manufacturing methods for future gearboxes so that they will not suffer the
same problems. Also, an accurate diagnosis of field problems will enable the user
to determine if a gearbox requires immediate modification or replacement or if
it can be expected to continue to function for some period of time, when repairs
can be made more conveniently.

Breakage

Breakage of gear teeth is the most catastrophic form of failure. It occurs precipi­
tously with no advance warning. If a number of teeth break, load transmission is
no longer possible. If only one tooth or a portion of a tooth breaks, there is the
possibility of secondary damage if the broken part interferes with other compo­
nents in the system. Also, dynamic loading will increase and if operation con­
tinues, other teeth will soon fail. Breakage of gear teeth is caused by excessive
bending stress in the root imposed by the transmitted load. Tooth breakage can
be the result of a fatigue mechanism (Figure 12.1) or an overload which exceeds
the gear tooth fracture strength (Figure 12.2). A fatigue break initiates as a small
crack which, over a large number of load cycles, propagates until a portion of,
or a whole tooth, separates from the gear. Failures of this nature may be a result
of system overloads greater than the design load, such as torsional vibrations.
Manufacturing discrepancies such as tool marks or metallurgical discrepancies
can also lead to fatigue failures by initiating cracks. In surface-hardened gears, if
the case in the root is not correct either due to heat-treat problems or excessive
machining, bending failures can result.
Maintenance and Failure Analysis 389

Figure 12.1 Fatigue breakage. (Courtesy of American Gear Manufacturers

Association, Arlington, Va.)

As shown in Figure 12.1, the smooth appearance of the fracture surface

attests to the fact that considerable working of the cracked surfaces occurred
prior to final separation. Overload breakage occurs in relatively few cycles;
therefore, the fracture surfaces are rougher than those of a high-cycle fatigue
failure. Some causes of overload breakage are:
Large particles passing through the mesh
Sudden misalignments such as when a coupling fails
Bearing seizures
Shock loads such as short circuits in a generator drive

Figure 12.2 Overload breakage. (Courtesy of American Gear Manufacturers

Association, Arlington, Va.)
390 Gear Drive Systems

In some cases other failure modes such as wear or pitting may increase dynamic
loading or weaken the tooth to the extent that breakage ultimately occurs.

Wear

Wear may be defined as the loss of metal due to the rubbing action of two
surfaces moving in relation to one another when the oil film is not of sufficient
thickness to separate them [1]. One form of rubbing wear is adhesive wear
characterized by metal particles from one gear tooth adhering to the mating gear
tooth by a welding action and subsequently detaching. The other form of
rubbing wear is abrasive wear caused by abrasive action between the sliding gear
teeth or by the presence of abrasive particles between them. These particles may
be contaminants in the oil or pieces detached from the tooth surfaces them­
selves. Figure 12.3 illustrates some worn areas on a helical gear.
Gear teeth do not necessarily wear during operation. If the oil film thick­
ness is sufficient to separate the mating tooth surfaces, millions of cycles can be
accumulated with no measurable wear. In some cases there may be an initial
wearing in of gear teeth and if rubbing wear diminishes with time it may not be
detrimental; therefore, when wear is first noted the gearset should be closely
monitored to determine the rate at which wear progresses such that a deter­
mination can be made as to whether it will be damaging. The wear-in

Figure 12.3 Tooth wear.

Maintenance and Failure Analysis 391

Figure 12.4 Initial pitting. (Courtesy of American Gear Manufacturers Associa­

tion, Arlington, Va.)

phenomenon is more common with through-hardened gears (Rc 32 approxi­

mately) than surface-hardened gearing (approximately Rc 60).

Pitting

Pitting manifests itself in several forms. Initial pitting, sometimes called frosting,
may occur during early operation. This is a surface-oriented failure mode where
local high spots contact and exhibit distress, as shown in Figure 12.4. As the
high spots are removed, the load is more evenly distributed and the pitting
action diminishes. This condition may be acceptable.
Destructive fatigue pitting is a result of repeated stress cycling of the tooth
surface beyond the material’s endurance limit. Surface or subsurface cracks
initiate, propagate, and eventually material detaches from the tooth surface,
leaving pitted areas (Figure 12.5). Pitting may progress to a point where large
areas are broken out, as shown in Figure 12.6. This condition is referred to as
spalling.

Scoring

Scoring is a form of surface damage on the tooth flank which occurs when over­
heating causes the lubricant film to become unstable, allowing metal-to-metal
contact. Local welding is initiated and the welded junctions are torn apart by
the relative motion of the gear teeth, resulting in radial score marks. Figures 12.7
to 12.9 illustrate degrees of the scoring phenomenon. Light scoring which does
not progress may be acceptable, but heavier scoring can destroy the tooth profile
and lead to pitting and breakage. The scoring type of failure mode generally
occurs in high-speed applications using low-viscosity lubricants.
392 Gear Drive Systems

Figure 12.5 Pitting.

Evaluation of Surface Distress

Tooth breakage is the only failure mode that demands immediate action be
taken. Wear, pitting, and scoring are progressive and may occur over a long
period of time without significantly impairing the function of the unit. As an
example, Figures 12.10 to 12.16 show the progressive deterioration of a gear
during 1000 hr of operation. At 50 hr there is a band of wear in the dedendum
of the teeth with a depth of 0.0001 in. at the right hand end increasing to
0.0006 in. at the more heavily worn left-hand end. Some pitting is also notice­
able. At 120 hr the wear has spread toward the tip of the tooth. More pits have
appeared. At 200 hr the tooth tip has worn further, to a depth of approximately
0.0006 in. At 400 hr the wear band has spread to the right-hand end. During
this segment of testing the right-hand end wore 0.0006 in. From 400 to 1000
hr there were no major changes in gear tooth condition.
The purpose of this discussion is not to advocate running gears in a worn
and pitted condition. The system noise and vibration level will increase as the
tooth surface deteriorates and abrasive particles are being introduced into the
lubrication system. There is always the danger of catastrophic breakage of a
weakened tooth. It can be seen, however, that even when the tooth surface is
severely deteriorated, gear teeth are capable of operating.
As a general guideline, if surface distress is noted early in operation and it
is light in nature, time can be taken to observe the progression and if distress
Gear Drive Systems
Maintenance and Failure Analysis 395

Figure 12.9 Heavy scoring.

ceases to increase, the gearset is probably satisfactory for continued operation. If

distress occurs after long operation, it is probably an indication that something
has changed in the operating conditions. Possibly loads have increased or the
lubricant has deteriorated. When surface distress appears there are several
potential changes in operating procedures that can be made to prolong gear life:
It may be possible to reduce the load on the gear teeth. Sources of external
loading such as coupling unbalance or misalignment should be checked. If
396
Gear Drive Systems

Figure 12.10 Tooth condition after 50 hr.

Maintenance and Failure Analysis 397

torsional vibrations are suspected, measurements should be taken. It may

be possible to derate the system and operate at lower loads.
An oil with higher load capacity may be available. EP additives can possibly be
of value. The manufacturer should be consulted before changing lubricants
Operating the unit with cooler oil which will be at a higher viscosity creating a
thicker film may be beneficial when surface distress is a problem.

Figure 12.13 Tooth condition after 400 hr.

 Gear Drive Systems

Figure 12.14 Tooth condition after 600 hr.

Figure 12.15 Tooth condition after 800 hr.

Maintenance and Failure Analysis 399

Figure 12.16 Tooth condition after 1000 hr.

Cracking
On occasion, cracks are observed on gear teeth as shown in Figure 12.17. Cracks
are usually a result of incorrect heat treatment or abusive grinding. In some cases
improper heat treatment results in brittle tooth tips or ends that chip off during
operation.
Cracks on the tooth flanks are dangerous since they can propagate and
lead to catastrophic failure; therefore, gears exhibiting cracking should be
replaced. Chipping of gear teeth at the tips or ends may be repairable by
blending. The danger here is that if more chipping occurs, the pieces may inter­
fere with rotating components.
Improper heat treatment of case-hardened gears can result in case-core
separation. Cracks originating in the core material propagate along the case-to-
core boundary and then work out to the surface. When this occurs large areas of
material are removed.

Oil Starvation
Quite a few gearboxes have been lost because the units were run without lubrica­
tion. Sometimes when viewing the components after failure it is not obvious
what the cause was. For instance, Figure 12.18 shows the condition of compo­
nents of a planetary gear set after 1.75 min of operation at full load and speed
without oil. The smaller sun gear has failed catastrophically, shedding all its
teeth. Other gears and antifriction bearings are in reasonably good condition.
The sun gear, having the least area to dissipate heat, reached the melting point
400 Gear Drive Systems

Figure 12.18 Oil starvation failure. (Courtesy of AVCO, Lycoming Corpora­

tion, Stratford, Conn.)
Maintenance and Failure Analysis 401

of its teeth and suffered the most damage. It was not obvious from the compo­
nent condition that oil starvation was the cause of failure. With journal bearing
gearboxes the diagnosis is easier since the bearings will sieze long before the
melting point of gear teeth is reached.

REFERENCE

1. Southwest Research Institue, Gear Tooth Scoring Investigation, Report

USAAMRDL-TR-75-33, San Antonio, Tex., July 1975.
APPENDIX
IN V O L U T E TABLES (IN V <j> = tan <p - <p)

Angle (deg) Cosine Involute Angle (deg) Cosine Involute

17.0 0.956305 0.009025 19.2 0.944376 0.013134
17.1 0.955793 0.009189 19.3 0.943801 0.013346
17.2 0.955278 0.009355 19.4 0.943223 0.013562
17.3 0.954761 0.009523 19.5 0.942641 0.013779
17.4 0.954240 0.009694 19.6 0.942057 0.013999
17.5 0.953717 0.009866 19.7 0.941471 0.014222
17.6 0.953191 0.010041 19.8 0.940881 0.014447
17.7 0.952661 0.010217 19.9 0.940288 0.014674
17.8 0.952129 0.010396 20.0 0.939693 0.014904
17.9 0.951594 0.010577 20.1 0.939094 0.015137
18.0 0.951057 0.010760 20.2 0.938493 0.015372
18.1 0.950516 0.010946 20.3 0.937889 0.015609
18.2 0.949972 0.011133 20.4 0.937282 0.015850
18.3 0.949425 0.011323 20.5 0.936672 0.016092
18.4 0.948876 0.011515 20.6 0.936060 0.016337
18.5 0.948324 0.011709 20.7 0.935444 0.016585
18.6 0.947768 0.011906 20.8 0.934826 0.016836
18.7 0.947210 0.012105 20.9 0.934204 0.017089
18.8 0.946649 0.012306 21.0 0.933580 0.017345
18.9 0.946085 0.012509 21.1 0.932954 0.017603
19.0 0.945519 0.012715 21.2 0.932324 0.017865
19.1 0.944949 0.012923 21.3 0.931691 0.018129

403
404 Gear Drive Systems

Angle (deg) Cosine Involute Angle (deg) Cosine Involute

21.4 0.931056 0.018395 25.3 0.904083 0.031130
21.5 0.930418 0.018665 25.4 0.903335 0.031521
21.6 0.929776 0.018937 25.5 0.902585 0.031917
21.7 0.929133 0.019212 25.6 0.901833 0.032315
21.8 0.928486 0.019490 25.7 0.901077 0.032718
21.9 0.927836 0.019770 25.8 0.900319 0.033124
22.0 0.927184 0.020054 25.9 0.899558 0.033534
22.1 0.026529 0.020340 26.0 0.898794 0.033947
22.2 0.925871 0.020629 26.1 0.898028 0.034364
22.3 0.925210 0.020921 26.2 0.897258 0.034785
22.4 0.924546 0.021216 26.3 0.896486 0.035209
22.5 0.923880 0.021514 26.4 0.895712 0.035637
22.6 0.923210 0.021815 26.5 0.894934 0.036069
22.7 0.922538 0.022119 26.6 0.894154 0.036505
22.8 0.921863 0.022426 26.7 0.893371 0.036945
22.9 0.921185 0.022736 26.8 0.892586 0.037388
23.0 0.920505 0.023049 26.9 0.891798 0.037835
23.1 0.919821 0.023365 27.0 0.891007 0.038287
23.2 0.919135 0.023684 27.1 0.890213 0.038742
23.3 0.918446 0.024006 27.2 0.889416 0.039201
20.4 0.917755 0.024332 27.3 0.888617 0.039664
20.5 0.917060 0.024660 27.4 0.887815 0.040131
20.6 0.916363 0.024992 27.5 0.887011 0.040602
20.7 0.915663 0.025326 27.6 0.886204 0.041076
20.8 0.914960 0.025664 27.7 0.885394 0.041556
20.9 0.914254 0.026005 27.8 0.884581 0.042039
24.0 0.913545 0.026350 27.9 0.883766 0.042526
24.1 0.912834 0.026697 28.0 0.882948 0.043017
24.2 0.912120 0.027048 28.1 0.882127 0.043513
24.3 0.911403 0.027402 28.2 0.881303 0.044012
24.4 0.910684 0.027760 28.3 0.880477 0.044516
24.5 0.909961 0.028121 28.4 0.879649 0.045024
24.6 0.909236 0.028485 28.5 0.878817 0.045537
24.7 0.908508 0.028852 28.6 0.877983 0.046054
24.8 0.907777 0.029223 28.7 0.877146 0.046575
24.9 0.907044 0.029598 28.8 0.876307 0.047100
25.0 0.906308 0.029975 28.9 0.875465 0.047630
25.1 0.905569 0.030357 29.0 0.874620 0.048164
25.2 0.904827 0.030741 29.1 0.873772 0.048702
Appendix 405

Angle (deg) Cosine Involute Angle (deg) Cosine Involute

29.2 0.872922 0.049245 33.1 0.838671 0.074188

29.3 0.872069 0.049792 33.2 0.836764 0.074932
29.4 0.871214 0.050344 33.3 0.835807 0.075683
29.5 0.870356 0.050901 33.4 0.834848 0.076439
29.6 0.869495 0.051462 33.5 0.833886 0.077200
29.7 0.868632 0.052027 33.6 0.832921 0.077968
29.8 0.867765 0.052597 33.7 0.831954 0.078741
29.9 0.866897 0.053172 33.8 0.830984 0.079520
30.0 0.866025 0.053751 33.9 0.830012 0.080305
30.1 0.865151 0.054336 34.0 0.829038 0.081097
30.2 0.864275 0.054924 34.1 0.828060 0.081894
30.3 0.863396 0.055518 34.2 0.827081 0.082697
30.4 0.862514 0.056116 34.3 0.826098 0.083506
30.5 0.861629 0.056720 34.4 0.825113 0.084321
30.6 0.860742 0.057328 34.5 0.824126 0.085142
30.7 0.859852 0.057940 34.6 0.823136 0.085970
30.8 0.858960 0.058558 34.7 0.822144 0.086804
30.9 0.858065 0.059181 34.8 0.821149 0.087644
31.0 0.857167 0.059809 34.9 0.820152 0.088490
31.1 0.856267 0.060441 35.0 0.819152 0.089342
31.2 0.855364 0.061079 35.1 0.818150 0.090201
31.3 0.854459 0.061721 35.2 0.817145 0.091066
31.4 0.853551 0.062369 35.3 0.816138 0.091938
31.5 0.852640 0.063022 35.4 0.815128 0.092816
31.6 0.851727 0.063680 35.5 0.814116 0.093701
31.7 0.850811 0.064343 35.6 0.813101 0.094592
31.8 0.849893 0.065012 35.7 0.812084 0.095490
31.9 0.848972 0.065685 35.8 0.811064 0.096395
32.0 0.848048 0.066364 35.9 0.810042 0.097306
32.1 0.847122 0.067048 36.0 0.809017 0.098224
32.2 0.846193 0.067738 36.1 0.807990 0.099149
32.3 0.845262 0.068432 36.2 0.806960 0.100080
32.4 0.844328 0.069133 36.3 0.805928 0.101019
32.5 0.843391 0.069838 36.4 0.804894 0.101964
32.6 0.842452 0.070549 36.5 0.803857 0.102916
32.7 0.841511 0.071266 36.6 0.802817 0.103875
32.8 0.840567 0.071988 36.7 0.801776 0.104841
32.9 0.839620 0.072716 36.8 0.800731 0.105814
33.0 0.838671 0.073449 36.9 0.799685 0.106795
INDEX

Accelerometers, 375-377 Antifriction Bearing Manufacturers

Acceptance tests, 336, 337 Association:
Alignment at installation, 331-334 dimensions and tolerances, 171
Allowable stresses: life ratings, 156
bending, 101, 119 Application factor:
compression, 105,118 bending, 99
splines, 142 compression, 105
Alloy steels, 249-252 Assembly of planetary gears,
Almen strip test, 279 312-317
American Gear Manufacturers Axial groove bearings, 188
Association: Axial pitch, 53
lubricants, 217
noise standard, 350
quality classes, 79 Babbitt, 197-199
rating standards, 93, 107-118 Backlash, 68
sound measurement standard, Back-to-back test rig, 339
350 Ball bearings, 147-153
vibration standard, 365 Ball thrust bearings, 152,153
American Petroleum Institute stan­ Base helix angle, 50
dards, 93,118,119 Base pitch, 28
Angular contact ball bearings, Base radius, 23
150 Bearings:
Annealing, 252 cost, 147, 181, 185,190
Antifriction bearings (see Rolling dimensions and tolerances, 171,
element bearings) 172

407
408 Index

[Bearings] Composite tooth inspection, 283-285

failures, 179-181,201-203 Compound planetary, 300, 301
heat generation, 177, 178 Compressive stress, 103-107
journal bearings, 185-204 failures, 391
life rating, 146, 156-170 Condition monitoring, 359-382
lubrication, 146, 176-178, 191, chip collection, 378-381
192 oil analysis, 361-363
materials, 195-199 oil pressure, 381, 382
mounting, 173-175 oil temperature, 381, 382
planet gears, 320-323 vibration, 363-377
selection example, 167-170 Contact ratio:
selection of configuration, axial, 53
145-147 internal mesh, 56-59
simultaneous radial and thrust profile, 42,45
loads, 166 total, 53
speed limitations, 146, 163 Coolers, 238
Bending stress, 96, 97 Core hardness, 255
failures, 388 Costs, 287-291
Breakage of teeth, 388 bearings, 181-185
Breathers, 240 housings, 259-260
Brinell hardness, 260-262 materials, 290, 291
planetary gears, 323
quantity effect, 291
Caliper measurement of tooth Couplings, 327-331
thickness, 67 Crowning, 84-87, 274
Carbides, 255 Cylindrical roller bearings, 154-156
Carbon face seals, 211, 212
Carburizing, 246, 249, 252-258
Case core separation, 399 Deaeration, 239
Case depth, 253, 254 Decarburization, 255
Case hardness, 255 Deep groove ball bearings, 148-150
Castings, 259 Degrees of roll, 42-46
Cavitation, 233 Diametral pitch, 26
Center distance, 26, 68-73 Diaphragm couplings, 329
Centrifugal force on planet gears, Differential planetary systems, 296
320-322 Disc couplings, 329
Chip collection, 378-381 Double helical gearing, 13
Circular pitch, 27, 53 Double row ball bearings, 152
Clearance of journal bearings, 191 DPH hardness, 261, 262
Coefficient of friction: Drawing format:
bearings, 177 gears, 76-87
gears, 129 splines, 89
Index 409

Duplex ball bearings, 150-152 Filtration:

Dwell time of lubricant, 239, 240 beta rating, 235
Dynamic factor: media, 236
bending, 99,100 micrometer rating, 234
compression, 105 Fittings, 243
Flash temperature index, 123-129
Forced-feed lubrication, 226-229
Economics, 287-291, 323 Forgings, 258, 259
Efficiency, 17, 228, 355-358 Form diameter, 26, 43
Elastic coefficient, 103 Foundations, 325-327
Elastic couplings, 330 Four-square test rig, 339
Elastomeric lip seals, 208-211 Frosting, 391
Elliptical bearings, 188 Functional inspection, 283-286
Environment, 19, 224
Epicyclic gearing, 293
Error in action, 285 Gearbox arrangements:
Extreme pressure lubricants, 219- concentric, 2
221 mounting, 2-6
parallel offset, 2
planetary, 10
Fabrications, 259 right angle, 7
Face contact ratio, 53 Gear couplings, 329
Face seals, 211, 212 Geometry factor:
Failure analysis: bending, 97-99
antifriction bearings, 179-181 compressive stress, 103-105
breakage, 388 Grinding, 274-276
cracking, 399 Grouting, 325, 326
fatigue, 388, 391
fracture, 389
journal bearings, 201-203 Hand of helix, 53
oil starvation, 299-301 Hardening:
pitting, 391 carburizing, 246, 247, 252-258
scoring, 391 hardenability, 249
spailing, 391 nitriding, 246, 258
surface distress, 392-399 through hardening, 245, 247, 252
wear, 390 Hardness, 260-262
Fatigue failures: Hardness ratio factor, 105, 106
bearings, 156-159 Heat generation, 355-358
gears, 388, 391 Heat treating:
Film thickness criterion: annealing, 252
bearings, 163 carburizing, 246, 247, 252-258
gears, 130-135 hardenability, 249
410 Index

[Heat treating] [Instrumentation]

nitriding, 246, 258 oil pressure, 381, 382
normalizing, 252 oil temperature, 381, 382
tempering, 247, 249 proximity probes, 365-371
Helical gears: sound,348-355
advantages, 13 speed, 346
double helical, 13 torque, 346, 347
geometry, 50-54 velocity pickups, 373, 374
lead, 51 Internal clearance of bearings, 171,
length of contact, 54 172
normal plane, 51 Internal gears:
Helical overlap, 53 contact ratio, 56-59
Helical splines, 91 geometry, 55
Helix angle, 51, 283 Involute:
High speed units: coordinates, 26, 30, 32
characteristics, 12 diagram, 84-86
power losses, 358 generation, 28
Hobbing: inspection, 279-281
generation of tooth form, 28 nomenclature, 26, 27
hob geometry, 29 properties of, 21, 23-26,47
manufacturing process, 267-270
Honing, 278
Hunting teeth, 318 Jet lubrication, 226-229
Jominy test, 250, 251
Journal bearings, 145, 185-204
Immersion heaters, 240 bearing parameter, 187
Inspection: clearance, 191
composite, 283-285 costs, 190
involute, 279-281 failure analysis, 201-203
lead, 283 lubrication, 191,192
spacing, 283 materials, 195-199
Installation, 325-334 pressure-fed configurations, 188-
alignment, 331-334 190
couplings, 327-331 salvage, 203
foundations, 325-327 speed limitation, 146
grouting, 325, 326 thrust bearings, 192-195
misalignment, 328 unit loading, 187,190,191
shimming, 325
Instrumentation:
accelerometers, 375-377 Keyways:
accuracy, 346 rating, 138
chip collection, 378-381 allowable stresses, 139
Index 411

K factor, 118 Lubrication:

Kinematic viscosity, 134 bearings, 146, 176-178, 191, 192
Knoop hardness, 261, 262 breathers, 240
components, 229-243
cooling, 238
Labyrinth seals, 205-208 filtration, 234-238
Lambda ratio: fittings, 243
bearings, 163 oil flow, 228
gears, 130 oil pressure, 229
Lapping, 271 piping, 241
L 10 Bearing life concept, 146, 156- reservoirs, 238-240
170 selection, 216-223
Lead: systems, 223-229
definition, 51 theory, 213
diagram, 86
inspection, 283
tolerance, 82, 87 Magnaflux, 262
Length of contact, 53, 54 Magnetic particle inspection, 262-
Life factor: 264
bending, 101 Maintenance, 383-385
compression, 106 condition monitoring, 359-382
Line of action, 41 Manufacturing, 267-280
Lip seals, 208-211 grinding, 274
Liquid penetrant inspection, 264- hobbing, 267-270
266 honing, 278
Load distribution factor: lapping, 271
bending, 101 shaping, 271
compression, 105 shaving, 273
Loads, gear tooth, 94-96, 299 shot peening, 278, 279
Load sharing in planetaries, 309-311 Materials:
Lubricants: alloy steels, 249-252
AGMA, 217 bearings, 162, 195-199
extreme pressure, 219-221 costs, 290, 291
kinematic viscosity, 134 drawing definition, 77
pour point, 216 elastomers, 210
pressure viscosity coefficient, 132 heat treatment, 247-258
rust- and oxidation-inhibited, 219 housings, 259, 260
SAE, 215 selection, 245-247
selection, 216-223 Measurement over balls, 59-67
specific gravity, 133, 214 Mesh frequency, 318
synthetic, 221-223 Metallurgy:
viscosity, 213 alloy steels, 249-252
viscosity index, 216 bearings, 162
412 Index

[Metallurgy] Oil:
carbides, 255 acid number, 363
gears, 15 analysis, 361-363
hardenability, 249 drain, 384
hardness, 260-262 flow, 228
heat treatment, 247-258 pressure, 228, 381, 382
material selection, 245-247 starvation, 399-401
retained austenite, 255 tanks, 238-240
steel quality, 251, 252 temperature, 381, 382
Minimum film thickness criterion: viscosity, 363
bearings, 163 water content, 363
calculation, 130-135 Overhaul, 385
lambda ratio, 130, 163
pressure viscosity coefficient,
132 Parallel offset, 331
Misalignment at assembly, 328 Pattern inspection, 286
Modification of tooth form: Piping, 241
lead, 83, 85-87, 274 Pitch line velocity, 13, 49
profile, 83,84-86, 274 Pitch tolerance, 82
Mounting, 325 Pitting, 391
flange, 4 Planetaries, 293-323
foot, 3 arrangements, 11, 295, 300-306
shaft, 4 assembly, 312-317
Multistage planetaries, 301-303 bearing loads, 320-323
compound planetaries, 300, 301
design considerations, 308-320
Needle bearings, 156 economics, 323
Nital etch, 265, 266 load sharing, 309-311
Nitriding, 246, 258 multistage, 301-303
Noise, 348-355 power feedback, 306-308
Nondestructive testing: rating, 309-311
liquid penetrant inspection, 264- ratios, 10, 295-308
266 reverted, 301, 302
magnetic particle inspection, 262- size comparison, 294, 295
264 split power, 304-306
temper etch, 265, 266 Stoeckicht, 311
Normalizing, 252 tooth loads, 299
Normal plane, 51 Pour point, 216
Power feedback systems, 306-308
Power rating:
Octave frequency bands, 351 bending, 102
Offset half bearings, 189 compression, 107
Index 413

Pressure angle: Reverted planetary, 301, 302

definition, 23 Rockwell hardness, 260-262
normal plane, 51 Roll angles, 24, 42-46
Pressure relief valve, 233 Roller bearings, 154-156
Pressure viscosity coefficient, 132 Rolling element bearings, 145, 147-
Profile contact ratio, 42, 45 156
Profile tolerances, 82 angular contact ball bearings, 150
Prorating bearing life, 170 ball thrust bearings, 152, 153
Protuberance, 276 cost, 181-185
Proximity probes, 365-371 cylindrical roller bearings, 154-156
Pumps, 229-234 deep groove ball bearings, 148-150
dimensions and tolerances, 171,
172
Quality class: double row ball bearings, 152
AGMA, 79 duplex ball bearings, 150-152
cost, effect on, 288-290 failures, 179-181
grinding, 274 heat generation, 177, 178
hobbing, 270 mounting, 173-175
shaping, 271 needle bearings, 156
shaving, 273 selection example, 167-170
self-aligning ball bearings, 152
simultaneous radial and thrust
Radius of curvature, 24 loads, 166
Rating: speed limitation, 146
criteria, 93 spherical roller bearings, 156
durability, 103-107 tapered roller bearings, 156
numerical example, 120-123 Rolling velocity, 47
planetaries, 309 Round edges, 78
standards, 93 Runout:
strength, 96-102 inspection, 283
thermal, 94, 223 tolerance, 79
Ratio, 6, 21 Rust- and oxidation-inhibited oils,
Recirculating power, 306-308 219
Redline inspection, 283
Reference surfaces, 76
Regenerative testing, 339 Salvage; journal bearings, 203
Reliability factor: Saybolt seconds, 214
bending, 102 Scoring:
compression, 107 analysis of high-speed gearset, 135
Resonance, 327 definition, 123-129
Retained austenite, 255 failure mode, 391
Reverse bending, 102, 103, 319 flash temperature index, 123
414 Index

[Scoring] Start up, 358, 359

probability, 128 Static tests, 343, 344
scoring criterion number, 129, 130 Steel:
Seals, 204-205 alloy steels, 249-252
carbon face seals, 211, 212 chemical composition, 246
elastomeric lip seals, 208-211 definition, 245
labyrinth, 205-208 hardenability, 249
Self-aligning ball bearings, 152 quality, 251, 252
Sequential mesh, 318 structure, 248
Service factors, 115-117 Stoeckicht, 311
Shafting: Storage, 385
arrangements, 2 Strength rating, 96-102
concentric, 2 Surface distress, 392-399
extensions, 2, 5 Surface finish, 76
key ways, 138 Surface hardening:
rating, 94, 136, 138 application, 245, 246
Shaper cutting: carburizing, 249, 252-258
manufacturing process, 271 drawing definition, 76, 77
tooth generation, 38 nitriding, 258
Shaving, 273 Synthetic oils, 221-223
Shot peening, 278, 279
Single flank tooth tester, 286
Size factor 100, 105 Tapered land bearings, 192
Sliding velocity, 47-49 Tapered roller bearings, 156
Sound testing, 348-355 Temperature factor, 102, 106
Spalling: Tempering, 247
bearings, 156, 179 Temper inspection, 265, 266
gears, 391 Test programs, 335-338
Spare parts, 385 acceptance tests, 336, 337
Specific gravity, 133, 214 efficiency, 355-358
Spectrographic oil analysis, 361-363 load testing, 337-343
Speed measurement, 346 regenerative testing, 339-343
Spherical roller bearings, 156 reports, 348
Spin tests, 336, 337 sound,348-355
Splash lubrication, 223-225 spin, 336, 337
Splines: static, 343, 344
allowable stresses, 142 test log, 337
ANSI standard, 88 Thermal rating:
design, 88-91 bearings, 177
rating, 139-142 gearboxes, 94, 223
Split power planetaries, 304-306 Through hardening, 245, 247, 252
Star gears, 298 Thrust washers, 192
Index 415

Tilting pad bearings, 190, 195 Troubleshooting, 385-388

Tolerances, tooth:
AGMA quality class, 79
lead, 82 Velocity pickup, 373, 374
pitch, 82 Vents, 240
profile, 82 Vibration, 363-377
runout, 79 accelerometers, 375-377
Tooth thickness: proximity probes, 365-371
backlash, 68 velocity pickups, 373, 374
calculation, 31, 53 Viscosity, 213, 223
caliper measurement, 67 Viscosity index, 216^
measurement over balls, 59-67
Tooth-to-tooth spacing, 82, 283
Wear, 390
Torque measurement, 346, 347
Total contact ratio, 53
Transverse plane, 51
Trochoid, 30, 33-41 Zyglo, 264