You are on page 1of 106


& ..=

Reproduced By G L O B A L
E N G t N E E R l N GD O C U M E N T S
c With T h e P e r m i s s i o n O f A G M A
U n d e r Royalty A g r e e m e n t



D e sig n G u idelines fo r A e r o s p a c e G e a rin g





(This In fo r m a tio n S h e e tis N O T a n A G M A S ta n d a r d )

AGMA 911-A94

AGMA 91%A94, Design Guidelines for Aerospace Gearing

CAUTION NOTICE: AGMA standards are subject to constant improvement, revision, or withdrawal as
dictated by experience. Any person who refers to any AGMA Technical Publication should be sure that the
publication is the latest available from the Association on the subject matter.
[Tables or other self-supporting sections may be quoted or extracted in their entirety. Credit lines should read:
Extracted from AGMA 911-A94, information Sheet - Design Guidelines for Aerospace Gearing, with the
permission of the publisher, the American Gear Manufacturers Association, 1500 King Street, Suite 201,
Alexandria, Virginia 223141.

This Information Sheet covers current gear design practices as they are applied to air vehicles and spacecraft.
The material included goes beyond the design of gear meshes and presents the broad spectrum of factors
which combine to produce a working gear system, whether it be a power transmission or special purpose
mechanism. Although a variety of gear types, such as wormgears, face gears and various proprietary tooth
forms are used in aerospace applications, this document covers only spur, helical, and bevel gears.

Copyright 0 1994 by American Gear Manufacturers Association

Published by
American Gear Manufacturers Association
1500 King Street, Suite 201, Alexandria, Virginia, 22314




AGMA 911-A94

Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..*...............


Scope .......................................................................


Application ...................................................................
References ...................................................................


Application ...................................................................


Definitions and symbols ........................................................

Definitions ...................................................................
Symbols .....................................................................



Designapproach ..............................................................
Design requirements and goals ... ;. ............................................
Identify design criieria .........................................................
. .......................
Preliminary design ....................................
Detail design



Lubrication ..................................................................
Cooling vs. lubrication requirements ............................................
Distribution systems ..........................................................
Lubrication system design considerations .......................................
Filtration ....................................................................
Oiipumps ...................................................................
Lube system condition monitoring ..............................................



Environmental issues .........................................................

Ambient temperature effects ...................................................
Ambient pressure effects ......................................................
Attitude effects ..............................................................
Contaminant effects (water, corrosives, dirt, dust, and sand) .......................
Vibration/Shock effects .......................................................
Fire resistance requirements ..................................................
Electromagnetic effects .......................................................
:. .......
Nuclear, biological, and chemical (NBC) effects .........................



Vibration and noise ...........................................................

Causes of gear vibration ......................................................
Consequences of vibration ....................................................
Design ......................................................................
Analyzing vibration problems ..................................................
VibratiorYNoise reduction techniques ...........................................



Load Capacity ...............................................................

Introduction ...............................................................
Spur, helical, and bevel gear tooth breakage and surface durability .................
Spur, helical, and bevel gear scuffing (scoring) -flash temperature index ............


Gear materials and heat treatment .............................................

Class and grade definitions ....................................................
Mechanical properties ........................................................

..3 9


AGMA 911-A94

Contents, continued


Cleanliness ...............................................................
Heat treatment ............................
Microstructure ...............................................................
Hardenability ................................................................
Dimensional stability ..........................................................
Pm-machining stock removal ..................................................
Ferrousgearing ..............................................................
Non-ferrous gearing ..........................................................
Material grades and heat treatment .............................................
Gear surface hardening .......................................................
Gear through hardening .......................................................

..4 8


Surface treatment ............................................................

Introduction .................................................................
Shot peening ................................................................
Ion implantation of gears ......................................................



Manufacturing considerations ..................................................

Introduction .................................................................
Spur and helical gears ........................................................
Bevel gears .................................................................
Stress relief treatment ........................................................



Gear inspection ..............................................................

General .....................................................................
Spur and helical involute gears ................................................
Bevelgears .................................................................



Rocket & space gearing .... ..................................................

Introduction .................................................................
Lubrication ..................................................................
Gear materials for space application ............................................


Symbols used in equations .....................................................


Aerospace lu&icant viscosities ................................................

Aerospace lubricant densities ..................................................
Aerospace lubricant pressure/viscosity coefficients ...............................
Aerospace lubricant specific heat values ........................................
Aerospacegreases ...........................................................
Aircraft dry lubricants .........................................................
Advantages & disadvantages of a common engine & transmission lubrication system .



Particle size distribution by weight ..............................................
Suggested functional test levels for propeller aircraft and turbine engine equipment ... 27
Suggested functional test peak levels for equipment installed on helicopters .........


Potential influence of design features on noise and vibration .......................

Vibration testing .............................................................
Sound and vibration reduction techniques .......................................




Tables, continued

Typical aerospace gear materials ..............................................



Surface coatings used in aerospacegear units .....................................



Candidate solid-film lubricants for space application ..............................

Candidate fluid lubricants for space application ..................................
Working fluid lubricants .......................................................


Retative life as function of lambda ...............................................
The general parallel-axis epicyclic gear train .....................................
Goodman diagram for combined load ...........................................


Typical aerospace lubrication system schematic .................................

Spur gear pump .............................................................
Vane pump ..................................................................
Gerotor Pump ...............................................................



Typical gearbox attitude limits ..................................................

Suggested vibration spectra for propeller aircraft and turbine engine equipment ......
Suggested vibration spectrum for equipment installed on helicopters ................
Terminal-peak sawtooth shock pulse configuration and its tolerance limits ...........



Sound and vibration paths .....................................................

Typical damping ring .........................................................



Diierent methods for determining tooth root stress ...............................

Directions of crack propagation in gear teeth ....................................
Reliabilii versus number of standard deviations .................................



Schematic of material ground from a gear tooth ..................................

Schematic of material ground from a distorted gear tooth ..........................



Fatigue strength in ground AISI 4349 (50 HRC) ..................................

Example of residual stress profile created by shot peening ........................
Peening 1045 steel at 48 HRC with 330 shot ....................................
Peening 1045 steel at 62 HRC with 330 shot ....................................
Stress profile of carburized gear tooth root, ground and then shot peened ...........
Increase in fatigue resistance of spiral bevel gear ................................
Fatigue tests on rear axle shafts ...............................................
Fatigue tests on notched shafts ................................................
Fatigue life comparison .......................................................
Correlation of Almen intensities as indicated by arc heights of peened strips .........



Heat treat coupon ............................................................



Spur gear geometry factor including internal meshes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Gearbox test and mission requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
References and bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

AGMA 911-A94

[The foreword, footnotes, and annexes are provided for informational purposes only and should not be
construed as a part of AGMA 911-A94, information Sheet - Design Guidelines for Aerospace Gearing,]
This Information Sheet supersedes AGMA Standard 411.02, Design Procedure forAirm&Engine
and Power
Take-OlT Spur and Helical Gears. Its purpose is to provide guidance to the practicing aerospace gear
engineer in the design, manufacture, inspection, and assembly of aerospace gearing. In addition, it
addresses the lubrication, environmental, and application conditions which impact the gearbox as a working
system of components.
Material in the Information Sheet is supplemental to current AGMA Standards, but does not constitute a
Standard itself. By definition, Standards reflect established industry practice. In contrast, some of the
practices discussed here have not seen enough usage to be considered standard, but they do provide insight
to design techniques used in stat-f-the-art
aerospace equipment. It is expected that the user of this
Information Sheet will have some general experience in gear and machine design, and some knowledge of
current shop and inspection practices.
Suggestions for the improvement of this information sheet will be welcome. They should be sent to the
American Gear Manufacturers Association, 1500 King Street, Suite 201, Alexandria, Virginia, 22314.


AGMA 911-A94

PERSONNEL of the AGMA Committee for Aerospace Gearing

Chairman: A. Meyer . . . . . . . . . . . . . . . Textron - Lycoming
Vice Chairman: K. Buyukataman . . . . Pratt & Whitney

J. Abrahamian . . . Pratt & Whitney
N. Anderson . . . . . GM Technical Center
I. Armitage . . . . . . Spar Aerospace
E. J. Bodensieck . Bodensieck Engineering
M. Brogiie . . . . . . . Dudley Technical Group
R.C. Bryant . . . . . . General Electric
Ft. Burdick . . . . . . . Aero Gear
J. Daly . . . . . . . . . . Metal Improvement Co.
R. Dayton . . . . . . . Wright Patterson A. F. 9.

R. Drago . . . . . . . . Boeing Helicopters

B. Dreher . . . . . . . Kaiser Aerospace
R. C. Ferguson . . Taiga Group
W. D. Glasow . . . . Sikorsky
T. Heiliger . . . . . . . Sikorsky
M. Howes . . . . . . . IIT Research
J. G. Kish . . . . . . . Sikorsky
E. A. Lake . . . . . . Wright Patterson A.F.B.
W. Michaels . . . . . Sundstrand

G. Belling . . . . . . . American Pfauter
J. D. Black . . . . . . General Motors
E. R. Braun . . . . . Eaton
C. E. Breneman . Advance Gear
A. T. Brunet . . . . . Allied Signal Aerospace
J. Cadisch . . . . . . Reishauer
H. S. Cheng . . . . . Academic Member
L. Cloutier . . . . . . Academic Member
9. Cluff . . . . . . . . . American Pfauter
F. W. Cumbow . . . M&M Precision
R. J. Cunningham Boeing
P. A. Deckowitz . . ITWlllitron
J. W. Dern . . . . . . SPECO Corporation
K. R. Dirks . . . . . . Allied-Signal, Garrett Eng. Div.
R. DiRusso . . . . . Kaman
D. W. Dudley* . . . Honorary Member
R. Durwin . . . . . . . Sikorsky
W. C. Emmerling Naval Air Propulsion Center
R. L. Errichello . . Academic Member
J. A. Ferrett . . . . . National Broach
D. J. Fessett . . . . Lucas Western IncJATD
H. K. Frint . . . . . . Sikorsky
R. Gefron . . . . . . . Superior Gear
N. L. Grace . . . . . Gleason Works
M. J. Gustafson . Kaman
D. R. Houser . . . . Academic Member
C. lsabelle . . . . . . Sikorsky
D. E. Kosal . . . . . National Broach
C. Layer . . . . . . . . Mmg
A. J. Lemanski
Academic Member
A. A. Lewis . . . : : Pratt & Whitney, Canada
M. Lonergan . . . . National Broach
P. Mangione . . . . Naval Air Warfare Center
W. Mark . . . . . . . . Academic Member

W. Marquadt * . . . Norwood Precision/Textron

D. Merritt . . . . . . . Lion Precision Gear
R. Miller . . . . . . . . Pratt &Whitney
J. Mogul* . . . . . . . Metal Improvement Co.
J. ODonnell . . . . . Naval Air Warfare Center
A. E. Phillips . . . . Emerson Power Transmission
T. L. Porter . . . . . . ITW/Spiroid
A. K. Rakhit* . . . . Solar Turbines
J. R. Reed . . . . . . Klingelnberg Soehne
T. Riley . . . . . . . . . NWL Control System
E. Ropac . . . . . . . Bachan Aerospace
S.S. Sachdev . . . Spar Aerospace
B. Schneider . . . . NASA, Johnson Space Center
D. J. Schreiner . . General Motors
A. Seireg . . . . . . . Academic Member
S. V. Shebelski . . Lion Precision Gear
E. E. Shipley . . . . Mechanical Technology
G. Skirtich . . . . . . Lion Precision Gear
L. J. Smith . . . . . . Invincible Gear
N. Sonti . . . . . . . . Academic Member
D. A. Sylvester . . Power-Tech
K. Tower . . . . . . . . Metal Improvement Company
D. P. Townsend* . NASA, Lewis
F. Uherek . . . . . . . Flender
M. Valori . . . . . . . . Naval Air Propulsion Center
L. Vesey . . . . . . . . iTW/Spiroid
D. A. Wagner . . . . General MotorsIAGT
H. Wagner . . . . . . Advance Gear
R. D. Wagner . . . National Broach
9. R. Walter . . . . . Liebherr Machine
R. F. Wasilewski . Arrow Gear
S. R. Winters . . . . General Motors
T. J. Witheford . . . Teledyne Vasco
G. I. Wyss . . . . . . Reishauer

* Contributed technical material to the document.


This page is intentionally blank.

. ..

AGMA 911-A94

AGMA 91%A94

Design Guidelines for

Aerospace Gearing
1 Scope
This Information Sheet covers current gear design
practices as they are applied to air vehicles and
spacecraft. The material included goes beyond the
design of gear meshes per se, and presents, for the
consideration of the designer, the broad spectrum of
factors which combine to produce a working gear
system, whether it be a power transmission or special purpose mechanism. Although avariety of gear
types, such as wormgears, face gears and various
proprietary tooth forms are used in aerospaceapplications, this document covers only conventional
spur, helical, and bevel gears.

1 .l Application
The working environment of the aerospace gear
has become so diverse that a single set of guidelines will no longer suffice. The operating conditions
imposed on a high speed, high powered, transmission or actuator are quite different than those experienced by the spacecraft mechanism which must
function in a hard vacuum for long periods of time
without maintenance. This Information Sheet addresses these differences and provides guidance to
the designer for these demanding applications.

1.2 References
The following standards contain provisions which,
through reference in thii text, constitute provisions
of this American Gear Manufacturers Information
Sheet. At the time of publication, the editions
indicated were valid. All standards are subject to
revision, and parties to agreements based on this
American Gear Manufacturers lnfomration Sheet
are encouraged to investigate the possibility of
applying the most recent editions of the standards
indicated below.
AGMA 230.01 - 1974, Surface Temper Inspection

AGMA 390.03a, - 1980, Gear Handbook - Gear

Classification, Materials and Measuring Methods
for Bevel, Hypoid, Fine Pitch Wormgearing and
Racks Only as Unassembled Gears.
ANSVAGMA 110.04- 1989, Nomenclature of Gear
Tooth Failure Modes.
ANSVAGMA 2000-A88, Gear Classification and
inspection Handbook - Tolerances and Measuring
Methods for Unassembled Spur and Helical Gears
(Including Metric Equivalents).
ANSVAGMA 2001-988,
Fundamental Rating
Factors and Calculation Methods for Involute Spur
and Helica/ Gear Teeth.
ANSVAGMA 2003-A86,
Rating the Pitting
Resistance and Bending Strength of Generated
straight Bevel, ZEROLB Bevel, and Spiral Bevel
Gear Teeth.
ANSVAGMA 20044389, Gear Materials and Heat
Treatment Manual
ANSVAGMA 6023-A88, Design Manual for
Enclosed Epicyclic Gear Drives.
ANSVAGMA 6123-A88, Design Manual for
Enclosed Epicyclic Metric Module Gear Drives.

2 Application
A listing of aerospace geared applications by type of
service or function performed is useful in segregating the diverse gearing tasks into mechanism families which experience similar load and environmental spectra.
Applications can be identified by general grouping
as follows:

- Main propulsion systems;

- Propeller gearboxes reduce


speed to propeller speed;

- Fan gearboxes allow the use of optimum

turbine and fan speeds for maximum efficiency;
- Helicopter transmissions. A system of
gearboxes and shafting to drive the helicopter
rotors from the engine(s);

AGMA 246.02A - 1983, Practice for Carburked

- Mechanical interconnection between

engines allow for independent engine opera-

Aerospace Gearing.

tion on multiingine


AGMA 911-A94

- Accessory drive gearboxes driie accessory

devices, such as generators, fuel pumps, hydraulic pumps, oil pressure and scavenge
pumps, blowers, alternators, etc;
- Auxiliary/secondary
power units (APU/
SPU) consist of an auxiliary turbine engine integrated with a gearbox to provide powerfor main
engine starting, electrical services, emergency
hydraulic power, cabin air conditioning, etc.;
- Actuators. A general class of geared devices
used to cause a position change of an object. The
objects may include aerodynamic control surfaces, winch cables, doors, landing gear, or
telerobotic arms. Actuators are distinguished
from most aerospace gearing in that they only
move on command;
- Space systems. A specialized grouping of
power (as in rocket turbo-pump drives), and actuatortype devices which have been designed to
be compatible with the unique rigors of outerspace environments. These include the high
power, short life rocket applications as well as the
long life satelliie or space platform systems.
3 Definitions

The terms used, wherever applicable, conform to

the following standards:
ANSI Y10.3-1968, Letter Symbok for Quantities
Used in Mechanics of Solids
AGMA 1012-FQO, Gear Nomenclature, Definitions
of Terms with Symbols
AGMA 904-689, Metric Usage
3.2 Symbols.
The symbols used in this information sheet are
shown in table 1.
NOTE - The symbols and definitions used in this
information sheet may differ from other AGMA
publications. The user should not assume that familiarsymbols can be used without a careful study
of these definitions.

SI (metric) units of measure are shown in parentheses in table 1 and in the text. Where equations require a different format or constant for use with SI
units, a second expression is shown after the first,
indented, in smaller type, and with M included in
the equation number.






w,Ka 1 4 KmKB

and symbols

K~K~ mF
The second expression uses SI units.

3.1 Definitions.

. ..(n)
. ..(llM)

Table 1 - Symbols used in equations




equation paragraph

Center distance

in (mm)


Application factor for pitting resistance





Surface condiiion factor for pitting resistance





Hardness ratio factor for pitting resistance





Life factor for pitting resistance




Load distribution factor for pitting resistance




Elastic coefficient




Reliability factor for pitting resistance



Size factor for pitting resistance



AGMA 91%A94

Table 1 (coMwo






Temperature factor for pitting resistance




Dynamic factor for pitting resistance





Lubricant specific heat

BTlJ/lbm F


&J&l K)

Pinion operating pitch diameter

in (mm)




Reduced modulus of elasticity



Face width

in (mm)



Effective or net face width

in (mm)



Oil film thickness

in (mm)



Heat generated at design point




Film thickness, minimum



Geometry factor for pitting resistance




Geometry factor for bending strength




Contact load factor for pitting resistance

lb/in* (MPa)



External application factor for bending strength

. - - - -




Rim thickness factor





Liie factor for bending strength





Load distribution factor for bending strength

- - - -




Reliability factor for bending strength




Size factor for bending strength





Temperature factor for bending strength





Dynamic factor for bending strength





Lengthwise curvature factor for bevel gear

- - - -



bending strength

Lubricant flow rate



Module ( = 25.4/pd )

11 M


Gear ratio (never less than 1 .O)



Number of standard deviations





Pinion speed



Transmitted power

hp NW)


AGMA 91 l-A94

Table 1 (concluded)



Diametral pitch ( = 25.4/m )






Reliabilii constant





Allowable contact stress number

lb/in* (MPa)




Allowable bending stress number

lb/in* (MPa)




Contact stress number

lb/in* (MPa)




Bending stress number

lb/in* (MPa)




Working contact stress number

lb/in* (MPa)




Working bending stress number

lb/in* (MPa)




Inlet oil temperature




Outlet oil temperature




Contact temperature

F (C)




Flash temperature





Bulk temperature



Speed parameter



Average rolling speed

inlsec (mm&c)



Entraining velocity


Load parameter

in/s (m/s)


Unit tangential load

lb/in (N/mm)



Normal unit load

lb/in (N/mm)



Tangential tooth load




Load sharing factor

lb 04


Pressure viscosity coefficient

in*/lb (l/MPa)


Specific film thickness

inlin (mm/mm)




reyns (kPa s)



Absolute viscosity


Coefficient of variation or standard deviation

reyns (kPa s)




Normal relative radius of curvature

in (mm)



Composite surface roughness

tin @ml



Surface roughness of pinion, gear

crin tw)


AGMA 911-A94

4 Design approach

4.1J Maintainability

4.1 Design requirements and goals

Guidelines for field service work, space requirements, and tool limitations must be specified early in
the project.

The design procedure begins with a definition of the

application, requirements, and goalsforthe project.
It is sometimes diiicuft to clearly define all aspects
of the project at the start, but a complete tabulation
of the following parameters should be made to
provide a working definition of the project.

4.1 .I Power/speed and torque/position

The complete range of power and speed or torque
and position (actuators) must be tabulated including
a definition of growth capabili.
A duty cycle
definition is required for calculation of life. Within
these parameters a design point must be selected
for sizing purposes.

4.1.2 Gear ratlo and direction of rotation

Gear ratio must be specified with an indication of
allowable deviation. Input and output directions of
rotation are required and are important in selection
of the hand of helix or hand of spiral for thrust
direction and lubrication considerations.

4.1.3 Life
A clear definition of required gear and bearing
system life must be provided. Life is defined at a
specified survival level.

4.1.4 Weight
System weight is criiical in aerospace applications.
A value for gear system weight should be specified
as dry gearbox weight or gearbox plus lubrication
system weight.

4.1.5 Size limita%ons

In most applications, gearbox location and maximum envelope will be defined. These details must
be made available to the designer.

4.1.6 Reliability
Reliability requirements are typically specified in
terms of mean time between failure (MTBF). A
historical data base of typical component reliability
will permit calculation of system reliability. New
products are more difficult to characterize. Tech-

nique& quantify reliability levels must be specified

for a new gearbox system.

4.1.6 Cost
Aerospace gearing is generally more costly than
commercial gearing because of the necessary
performance, qualii and traceability requirements.
Life cycle cost is often established at the start of the
project as a goal or as a requirement. Life cycle cost
is defined as the total cost of ownership of a system
over its operating fife.

4.1.9 Efficiency
In most aerospace applications, gearbox efficiency
is an important design consideration because it
influences system weight and power requirements.
Efficiency requirements and goals will provide the
designer a clear indication of the project objectives
and may affect key decisions in the design process.

4.1.I0 Altitude/attitude requirements

Altitude and attitude specifications are required for
lubrication system design, since oil pump and oil
passage design are dependent on these parameters. In lieu of any specific application data
MIL-E-3!593C provides general requirements for
aerospace applications.

4.1.I1 Externally generated gearbox loads

External loads can be generated by rotor loads,
flight maneuvers, gravity and gyroscopic effects,
hard or crash landing requirements, or vibration, as
applicable. All must be considered in the design of
the gearbox housing, mounts and their effects on
misalignment of bearings and gears within the
gearbox. Typical loads are given in MIL-E-3593C.

4.1.I2 Mount locations

Mount locations must be specified to allow design
and analysis of the housing and internal structure
under external loading conditions. Mount location
requirements may also affect maintainability considerations.

4.1 .I3 Loss of lubricant

All military and some commercial aircraft have
requirements for operation with loss of lubricant ,
typically specifying a time and power level of
operation after loss of lubricant. These require-

AGMA 911-A94

ments must be known to allow the design of a

suitable lubrication system.
4.1 .I4 Test requirements
Test requirements are sometimes different than
those used to design the gearbox. If an unusual test
is required it can affect the gearbox design.
4.1.15 Noise requirements
The recent trend in air vehicle specification has
been to require meeting specified internal noise
levels in cabin and cockpit.
4.2 identify design criteria
It is sometimes difficult to clearly define design
objectives or goals of a gearbox or gearset. Proper
identification of design criteria requires application
of many disciplines such as elastohydrodynamics,
involutometry, geometry, stress analysis, system
dynamics, materials, kinematics, vibration, heat
transfer, processes, manufacturing, economics,
etc. Each of the above disciplines requires that
design limits be imposed such as:
- Stress limits;
- Scuffing (scoring);
- Minimum oil film thickness;
- Type of mounts, deflections and locations;
- Weight and Cost:
- Vibration;
- Noise.
The design criteria which have the largest influence
on the final configuration are as follows.
4.2.1 Allowable contact stress
The tooth contact (Hertz) stress limit depends on
the type of application, required service life, properties of materials used, and the shape of the tooth
surfaces near the point of contact before the load
transfer begins.

as aircraft flap drive systems, winches, and spacecraft robotic manipulator arms. These loads are the
highest loads specified for the gears, and are often
two to three times higher than the maximum
continuous operating loads. This is particularly true
for low speed actuator gearing where there are no
significant dynamic loads. To properly accommodate these conditions, the designer must evaluate
the gear design for maximum compressivestresses
at the maximum holding loads.
Holding loads are usually specified as limit loads,
where there may be no permanent deformation or
yielding allowed, and ultimate loads, where deformation is allowed but the gears may not fracture.
A value of 3.1 times the shear yield strength may be
used as the allowable contact stress for most steels.
High strength, through hardened stainless gears,
may also be utilized where environmental conditions warrant. The surface durability of these gears
may be improved, if required, by nitriding.
4.2.2 Allowable bending stress
The allowable tooth root bending stress is a function
of the hardness and residual stress near the surface
of the root fillet and at the core. Power transmission
Power transmission gears are usually case hardened by either nitriding or carburizing to obtain
adequate high cycle bending and contact fatigue
A method for calculation of bending stress, along
with allowable limits, is given in ANSVAGMA
2001-B88. Actuator gearing Actuator gearing

Gears which are manufactured from high strength

through hardening steels (260 ksi and above), and
heat treated to through hardness in the Rockwell C
50+ range, have shown higher bending fatigue
strength in the lower fatigue cycle range (i.e. less
than lo6 tooth bending cycles) than conventional
case hardened gears. Thus, a designer seeking
optimum minimum weight gearing should consider
the actual cycle life imposed prior to making a
selection of either case hardened or high strength
through hardened gears for a particular application.

Actuator gears are subject to holdingloads which

Allowable bending fatigue limits are given in ANSI/

are static loads. These loads occur in systems such

AGMA 2001-B88.

4.2.1 .I Power transmission

In high pitch linevelocity gearsets, thedistribution of
dynamic load is required for accurate determination
of tooth contact stress. A method for calculation of
contact stresses, along with allowable limits, is
given in ANSI/AGMA 2001-888.

AGMA 911-A94

4.2.3 Surface temperature

The mechanism of surface failure due to a sudden
temperature rise is one of the major considerations
in aircraft gearing.
Each oil has a characteristic criiical temperature independent of gear design and operating conditions.
Appendix A of ANSVAGMA 2001-888 defines
scuffing as related to the instantaneous temperature rise on tooth surfaces caused by frictional heat.
The equations which define the surface temperature rise have begun to adapt dynamic conditions
and have become more representative of what
happens at the gear mesh, including: constrained
heat source on the tooth profile, sliding velocity
variations, tooth surface conditions, load sharing, oil
jet cooling, oil jet impingement depth and air/oil mist
Experiments have verified that minimum values of
surface temperature occur at operating pitch
A method of calculating surface
temperature is presented in Appendix A of
Maximum values
ANSVAGMA 2001-B88.
generally occur at or near the highest point of single
tooth contact. Although the above procedure is
currently in use, the method is only applicable under
boundary lubrication conditions. Allowable scuffing
temperature values should be based on the
lubricant temperature at which lubricant breakdown
occurs, the material tempering temperature, or the
users experience whenever possible.

is unit tangential load, lb/in (N/mm).

The currently used lubricant film thickness analysis

is the extension of a bearing film thickness study by
Osborne Reynolds. Ertel, Gruben, Hamrock, Dowson, and Higginson contributed to the equation in its
current form.
The most influential parameter in the calculation of
film thickness is the speed parameter U, which
represents the average rolling speed and the
surface condition of the point at which the EHD film
thickness is calculated.
Surface geometry and finish are important to the
EHD lubrication process. EHD theory is based on
the assumption of perfectly smooth surfaces, that is,
no interaction of surface asperities. In reality, this is
not true for boundary lubrication. Therefore, the
relative life chart was introduced.


. ..(2)

(31 and 02 are the roughnesses of the two surfaces

in contact and his the ratio of EHD film thickness to
composite surface roughness. A plot of h vs.
relative life is shown in figure 1. This figure assumes
sufficient loading and otherwise satisfactory operation of the gears.
NOTE-& issupplantingrms as a way of describing roughness. Both terms are still in use but are
not equivalent.

4.2.4 Lubricant film thickness

Lubricant film thickness has received ever-increasing attention since the time it was introduced by
Martin in London Engineering in 1914.


H = ,,,,g:

Aerospace gears



, A Bkaririgs



H is oil film thickness, in (mm);

p is viscosity, reyns (kPa s);
U is average rolling speed, inls (mm/s);


is normal relative radius of curvature, in



ecific film thickness, h

h < 0.4 Danger of scuffing for carburized gears

li 5 0.4 Acceptable, assuming boundary layer
Figure 1 - Relative life as a functio~~~&n@&

AGMA 911-A94

Further studies by NASA simplified the general

equation to the form presented in Appendix A of
ANSVAGMA 2001-B88 for dimensionless minimum film thickness:

H mh = 2.65



where the following are dimensionless parameters:

materials parameter, G;
. ..(5)
speed parameter, V;

- PO

42.5 Structural integrity

Structural integrity is achieved by the proper
definition of gear, bearing and gearbox mounts;
gear configuration and materials: selection of
bearings; type of bearings and bearing location;
seals and type of sealing surfaces.
4.3 Preliminary design


The areas of concern during the preliminary phase

of aerospace gearbox design consist primarily of
performance, cost, configuration and packaging.


load parameter, W,

45.1 Configuration study

is pressure-viscosity coefficient, in2/lb



is absolute viscosity, reyns (kPa s);


is entraining velocity, in/s (m/s);


is reduced modulus of elasticity, lb/in2



normal relative radiusof curvature, in (mm);

In the preliminary design stage, it is generally

necessary to lay out various gearbox configurations
which meet the basic speed, power, and ratio
requirements. These configurations can be compared against design requirements and rated
against each other in terms of reliability, efficiency,
maintainability, cost, size, weight, and similarity to
past experience. From this process the most
suitable configuration for the particular application
is selected.


load sharing factor; Gearing


normal unit load, lb/in (N/mm).

A large number of gearbox configurations are

possible to achieve the desired design goal, some
of which are described below. The gearbox
envelope is generally set by the space available
plus the speed, power and ratio requirements.
However, the configuration may be further complicated by pitch change mechanisms, accessories,
overrunning clutches, engine air intake, etc.
Possible configurations include:

The following enhancements may be added to the

calculation as follows:
- Transient squeeze film effects from change in
entrainment velocity, surface geometry and dynamic load;
-Actual dynamic&ad profile in place of average
tangential load;
- Equilibrium surface temperature and oil inlet
temperature which defines the temperature of
the oil film;
- Use of optimal, experimental heat transfer coefficients when oil jet cooling is used for minimization of surface temperature;
- Effects of oil entrapment on long face width
gears may be included and equations may be
separated from short face width gears.
The relative film thickness, as calculated using
equation 4 for ZYZ~ has been derived and used
successfully using narrow face width gears such as

accessory gears. If equation 4 is used for power

gearing without the previously noted enhancements, the definition of when boundary lubrication
occurs may be as low as R = 0.2 to E.= 0.4.

- Offset. This refers to a gearbox in which the

input and output shafts have a parallel offset;
- Inline. This refers to a gearbox axis in which
the input shaft and output shafts are concentric;
-Angular. This refers to a gearbox in which the
input and output shaft are at an angle to each
other. Epicyclic
In the same sense that some gearformsare specific
cases of a more general configuration (example: A
spur gear is the special case of a helical gear with a

AGMA 911-A94

zero helix angle), a gear system can be general or

specific. In the context discussed here, we will
consider the parallel axis epicyclic rather than the
more general bevel epicyclic. Refer to ANSVAGMA
Standard 6028-A88 or 6128-A88 Metric.
Kinematically, the general case for the parallel axis
epicyclic is an arrangement of six gears in two
planes as shown in figure 2. By definition a sun gear
is a gear element whose axis is coincident with the
system axis. Thus, the system shown contains four
sun gears; i.e., two external suns and two internal
suns. Internal sun gears are sometimes called ring
gears. The sun gears of each plane are meshed
with an idler. If the two idlers are assumed to be
mounted on a common shaft which is, in turn,
supported by bearings to a rotatable structure we
have the general parallel axis epicyclic system.
By controlling the location of the instant center of
rotation in the above system of gears, the designer
can produce 88 epicyclic variations, each with its
own unique properties.
Some of the more important variations have been
given names and appear in countless transmission
systems. For example:
- The simple epicyclic:
If each of the
corresponding gears in the general system are
assigned identical tooth counts, then the gearing
in one of the planes becomes redundant and may
be eliminated, leaving a single external sun, a
single internal sun, each meshed with a common
idler which is finally supported by the rotatable
structure usually called a carrier
In the general simple epicyclic, everything in theory
can rotate. However, by controlling the location of
the instant center of rotation, we can produce some
very interesting and important gear systems. These
- The simple planetary: If we constrain the internal sun against rotation its pitch circle has zero
angular velocity and the remaining three components, the external sun, the idler, and the carrier
are free to rotate. As the idler rolls in mesh with
the fixed internal sun it orbits about the system
axis as it rotates about its own axis, thus the idler
in a simple planetary has come to be called a
planer. The use of a single planet would place
serious balance constraints on the gear system,
so it is common practice to fit the carrier with mul-

tiple, equally spaced planets to assure a balanced system, and most importantly, provide
multiple load paths for reduced weight.


Figure 2 - The general parallel-axis epicyclic
gear train
If the input to the simple planetary is to the external
sun gear, the resulting gear box will be a speed
reducer, and conversely if the input is to the carrier,
the resulting gearbox will be a speed increaser. In
application the practical usable reduction ratio will
lie between 2.5 and 7 and the input and output
shafts will have the same direction of rotation.
- The star gearbox: If we constrain the carrier
against rotation, the system instant center of
rotation is coincident with the axis of the idler and
the rotating components become the central
external sun, the idler, and the internal sun.
Since the idler no longer orbiis about the system
axis it is usually called a star. Again, for reasons
of equilibrium and load division it is common
practice to fit the stationary carrier with multiple,
equally spaced stars.
If the input to the star gearbox is to the central
external sun, the resulting unit will be a speed
reducer, and conversely if the input is to the internal
sun, the resulting unit will be a speed increaser. In
application the practical usable reduction ratio lies
between 1.5 and 6 and the input and output shafts
will have opposite directions of rotation.
The star gear system has found extensive use in the
first reduction of high speed systems because it is

AGMA 911-A94

free from high centrifugal bearing loading caused by

orbiting planets.
- The solar gearbox:
If we constrain the
external sun against rotation the system instant
center of rotation is coincident with the pitch
circle of the external sun, and the rotating
components become the internal sun gear, the
planet, and the carrier. Since, in this system, all
components orbit about the central fiied member the name solar is quite descriptive.
Of the simple epicyclics described so far, the solar
system is the least popular since for a given
reduction ratio it has higher mesh velocities and a
lower transmission efficiency. Usable ratios lie in a
narrow band between 1.14 and 1.5 with driving and
driven shafts rotating in the same direction.
- The compound epicyclic:
Referring once
again to figure 2, if the tooth counts of the gear
elements on each end of the idler shaft are not
the same, then all elements in the system can be
relevant to the creation of useful gear arrangements. A few of the possible arrangements are
noteworthy and will be discussed further:
- The compound planetary. If either of the
internal suns is constrained against rotation its
pitch circle has zero angular velocity and the
remaining four components are free to rotate;
i.e., the two external suns, the compound planet,
and the other internal ring gear. In theory, the
designer could produce a transmission with
three output shafts, but it would be a rare system
where such a configuration would be useful.
There are numerous examples of flight systems
with counter rotating propellers which use the
concept of a compound planetary withtwooutput
shafts. As with the simple epicyclics, it is usual
practice to configure the gearbox with multiple
equally spaced planets to assure a balanced
drive, and multiple load paths.
In space robotic systems, extensive use is made of
the compound planetary using a single driving
external sun, one fixed internal sun and one output
internal sun. In this latter case, the carrier and the
second external sun of the general arrangement are
not utilized, and are therefore discarded.
Usable ratios available from the compound plane
tary cover a very wide range and can be found as
low as 5 to values well over 1000. The user is


cautioned however, that some compound planetary

variations exhibit very poor transmission efficiency
due to high effective pitch line velocity in the high
torque meshes. A thorough analysis of each
application is recommended before committing the
design to detailing. The parallel axis differential
This special case of the parallel axis epicyclic will be
mentioned here because of its extensive use in
spacecraft and other systems that require a redundant drive source. In such a system, use is made of
two suns, and two planet pairs. Each planet pair is in
mesh, and the first planet of each pair is in mesh
with one of the sun gears while the second planet of
each pair is in mesh with the other sun gear. The
carrier is free to rotate and is usually assigned to be
the output member. A motor/brake combination is
fitted to each of the input suns. In service, either of
the motors can be the system input, and the
opposite brake can serve as the system reaction
member. The reduction ratio of the differential is 2. Accessory drive system
The accessory drive system is a drive train dedicated to drive accessories (i.e., lube and scavenge
pumps, alternators, generators, etc.) which are
requirements of the application. The size and
location of the gearbox are dependent on the
accessory requirement, positioning of these accessories and the position of the gearbox input drive.
When positioning the accessory gearbox, consideration needs to be given to the overall configuration to ensure that a compact package is obtained.
Definition of an accessory drive system depends on
the spaces and the location available to driie the
accessories. One concern is the selection of gear
and bearing diameters to fill the distance between
the power input and available accessory mount
locations. Another concern is to ensure that system
life is compatible with the general requirements.
Both concerns are equally essential for a successful
drive train.
Refer to ANSVAGMA 6123-A88

for specific

45.2 Gear sizes

There usually are two modes of operation which
size gears as follows:
- Start up conditions;

AGMA 911-A94

- Spectrum of speed and torque or torque and

position operating conditions.
Gear tooth geometry definition plays an important
role in optimization of design. In general, proper
selection of tooth proportion, pressure angle, and
spiral angle or helix angle are important to increase
the overall contact ratio and to provide better
balance between operational stresses. Most recent
experience shows increased use of high transverse
contact ratio spur and helical gears.

apply to a given material and number of cycles of

operation, as shown in the note.


Case carburized
AMS6265 Material

4.3.3 Loads
Steady stress Dynamic loads

High power density gearing depends on designing
gears to maximum load capacity. For high speed
gears, a major requirement becomes the ability to
accurately calculate dynamic loads so the essential
performance and design characteristics can be
accurately predicted.
Dynamic load consists of three parts. The first part
is defined by the component or system resonance.
The second part is defined by the gear tooth mesh
transmission error, and the third part consists of
inputtorquefluctuation. Resonancecondiiions may
be controlled by changing gear and web configuration, by damping, or by changing gear ratio. The
gear tooth mesh generated conditions can be
controlled by changing the tooth form, contact ratio,
and damping. Additional mass elastic analysis can
also be performed to ascertain torsional dynamics. Centrifugal loads
Centrifugal Loads in gearboxes result from component rotation. These loads impose steady stresses,
for a given speed, on components such as gears,
bearings, and gearbox mounts. The stresses are
accounted for under combined loading. Combined loads (Goodman diagram)
For high speed gears the steady stress (centrifugal)
should be combined with the alternating stress at
the gear tooth root, as shown on a Goodman
diagram in figure 3. Combined operating stresses
such as point A fall within the area bounded by the
endurance limit and yield strength and are acceptable. The same gear operating at a higher speed,
point B, might fail since the combined stress
exceeds the allowable limit. Goodman diagrams

Figure 3 - Goodman diagram for combined

4.3.4 Rotating components
Preliminary design is not complete unless all
rotating components such as splines, shafts, bearings and seals are defined along with the gears. Splines
Splines are used to transmit torque between two
parts with a common axis. In a gearbox, splines
transmit torque between a shaft and a gear or a
shaft and another shaft. In practice there are
straight sided, serrated, and involute splines but in
aerospace transmissions, involute splines are normally used. Involute splines transmit torque through
contact between external and internal spline members independent of the fii clearance. This high
degree of contact reduces the wear and the length
of spline required. The mating internal and external
involutes provide a centering effect and the distribution of force from top to bottom is also very good.
Requirements for involute splines are usually specified in one of the following standards:
-ANSI B92.1, Involute Splines and Inspection;
-ANSI B92.2M, Metric Module /nvo/ute Splines
and Inspection;
- IS0 4156, Straight Cyl~ndricallnvolute Splines.
Involute splines can be of the side bearing, major
diameter, or minor diameter fii type. In the side
bearing fit types of spline, which is the most widely
used type, the mating members contact only on the
driving sides of the teeth with clearance between
the major and minor diameters. In the major
diameter fit type of spline, the mating members


AGMA 911-A94

contact and are piloted by the major diameters with

clearance on the minor diameters. Minor diameter
fit splines are only used in situations where the
diameter is too small for the cutter of the internal

accessory drives are sometimes designed with

non-metallic muff inserts between spline members.
These serve as an inexpensive compliant part
which mitigates metallic spline wear.

The splines can be designed to act as fiied,

non-working types or flexible, working types. In the
fixed spline, the members are piloted on one or both
ends, so that the pilots, rather than the spline teeth,
carry any radial load. The fiied type of splined joint
is often clamped in the axial direction. The objective
in the fiied spline design is to force the spline to
carry only torque while other elements carry radial
and axial load. Fixed splines must have clearance
because of non-concentricity between the spline
pitch diameter and the mounting diameters.
Without clearance, the internal and external members could bind, leading to increased operating

Bearings used in aerospace applications generally

are one of the following types:
- Deep groove ball bearings;
- Cylindrical roller bearings;
- Needle bearings;
- Angular contact ball bearings;
- Angular contact ball bearings with split
inner race;
- Tapered roller bearings;
-Journal bearings;
- Thrust bearings:
- Duplex bearings.

A flexible spline is not held radially by a diametral fit.

This permits both radial and angular misalignments
of the mating members. There is generally no axial
clamping in a flexible spline since this would tend to
restrain angular or radial motion. The spline should
have enough clearance to allow it to move in a
misaligned condition without binding. Splines which
must accommodate excessive misalignment
should be crowned along the flank to prevent end
loading and keep the load toward the center of the
tooth. Outside diameter crowning is also used to
ensure adequate root clearance under misaligned
A spline subject to angular misalignment carries an
induced bending moment across mating members
because friction at the spline teeth does not permit
free angular motion. The magnitude of the induced
moment is a function of torque, coefficient of friction,
angular misalignment, and component bending
Lubrication is beneficial to fixed splines and is
recommended for flexible splines, especially at high
speeds where the teeth tend to have more sliding
and wear. Filtered oil supplied to the spline joint
provides cooling and also washes away abrasive
particles. Grease packed splines are also used.
However, they tend to trap the abrasive particles, Bearings

As the bearing size increases, it is generally more

difficult to obtain calculated life due to changes in
preload caused by mounting, thermal and centrifugal load variations and deflections. Seals
The gearbox design is required to minimize the
number of static oil or grease seals to prevent
lubricant loss. Experience has shown that the use of
flat gaskets as static seals has been so poor that
they should be used only if absolutely necessary.
O-ring seals are generally used.
The dynamic seals can either be spring or magnetically loaded face seals, bore rubbing seals, labyrinth seals, or lip seals. Efforts should be made to
positively drain, and to provide pressure balance
and damping for any dynamic seal system.
Consideration should be given to the surface finish
and lay of shafts and journals which have contact
with seals. Either too fine or too coarse a surface
finish could be detrimental.
4.3.5 Lube system requirements
Details of the lube systems are discussed in clause
5. Consideration should be given to cool, lubricate
and scavenge all rotating power transmission
4.4 Detail design

which can accelerate wear and thus will require

Detail design of a geared system requires accurate

periodic maintenance. Flexible splines used as

evaluation of dynamic gear tooth loads caused by


AGMA 9ll-A94

load transfer from one mesh to another and

momentary overloads caused by system resonance. In detail design, structural gear analysis
requires an assessment of tooth load capacity, to
select or calculate derating factors. The design
process may be based on conventional AGMA or
FE analysis.
Manufacturing tolerances, tooth errors, profile
modifications and system misalignment will significantly influence gear tooth load along the contact
path, thus affecting load sharing.
Accurate evaluation of gear tooth load sharing
behavior under dynamic conditions is not only
important in minimizing the weight of the entire
system but also is valuable to enhance over- all
system reliability.
Detail design of aircraft gears can also involve
modifications of analysis methods, using nonlinear
multibody dynamic analysis including equilibrium
analysis, kinematic analysis, vibratory analysis with
open loop systems, closed loop systems and elastic
(flexible) and/or rigid body systems.
All of the above can be used to perform an
assessment of the load distribution along the
contact line. ANSVAGMA 20014388 defines load
distribution for gears of general use.
In addition to materials and design configurations,
the following items greatly influence the rate of load
transfer, or a systems response to input torque:
- Geometry of Pinion and Gear Teeth;
- Thermal Distortions;
- Gear Rim Centrifugal Forces;
- Profile Modifications and Crowning;
- Manufacturing and Alignment Errors;
- Instantaneous Angular Position of Gears;
- Rotational Delay of Driven to Driving Gear
(Angular Acceleration);
- Total Tooth Deflections (Rim, Web, etc.);
- Shaft Deflections (Bearing, Housing, etc.).
Load distribution is influenced by the above factors
and is non-uniform along the contact lines of
meshing gears. To determine tooth load distribution, tooth and rim deflections are required. These
deflections vary with the load position and affect the

4.4.1 Finite element modeling considerations

Single flank element models can be used to
determine tooth stress. To develop a finite element
methodology and a design tool to analyze the load
sharing behavior from simple spur gear systems to
more complex helical and spiral bevel gears on
combined systems, an attempt should be made to
address the factors influencing load sharing discussed earlier.
4.42 Tooth bending and contact stress considerations
Once the load distribution along the contact path is
obtained, the calculated load can be transferred to
gear tooth pair mesh locations to obtain stresses at
the root or along the contact surfaces. The calculations and limits are discussed in clause 8.
Gear stresses are a valuable design tool in determining thesize of the gears, and thus minimizing the
gear system weight. It is particularly important in
sizing (where possible) to base the selection of
derating factors of a new design on old designs
which are similar and have been successful in the
The tendency of gear teeth to pit has traditionally
been thought of as a surface fatigue problem in
which the prime variables are the compressive
stress at the surface, the number of repetitions of
the load, and the endurance strength of the gear
material. In steel gears the surface endurance
strength is quite closely related to hardness, so
stress, cycles, and hardness become the key items.
Gear work in the 1970s led to two very important
- Pitting isvery much affected by lubrication conditions;
- There is no pitting endurance limit. (S-N
diagram does not become asymptotic.) The
allowable stress used for design purposes considers such items as the number of cycles and
the types of material and oil used.
Work on the theory of EHD showed that gears and
rolling-element bearings often developed a very
thin oil film that tended to separate the two
contacting surfaces so that there was little or no

dynamics and tooth root stress as the tooth rotates

metal-to-metal contact.

through the entire mesh.

situation was obtained, the gear or the bearing

When this favorable


AGMA Qll-A94

could either carry more load without pitting or run for

a longer time without pitting at a given load.

finish, and designed to controlled surface finish and


Gears in service frequently run for several thousand

hours before pitting starts, or becomes serious. A
gear can often run for up to a billion (1Og)cycles with
little or no pitting, but after 2 or 3 billion (2 or 3 x 1Og)
cycles, pitting, and the wear resulting from pitting,
can make the gears unfit for further service.

Aircraft bearings are typically AFBMA grade 5 to 7

or better, selectively designed to meet performance

4.4.3 Regimes of lubrication

To handle the problem of EHD lubrication effects,
three regimes of lubrication should be considered
(see figure 2.12 in [19]*). These are:
- Regime I: No appreciable EHD oil film
- Regime II: Partial EHD oil film (mixed);
- Regime Ill: Full EHD oil film (full film).
Regime I is encountered in aircraft gears when
speeds are jaw, such as in the final stages of
gearing in a helicopter gearbox.
Regime II is characterized by partial metal to metal
contact. The asperities of the tooth surfaces hit
each other, but substantial areas are separated by a
thin film. Regime II is typical of medium speed
gears, highly loaded, running with a relatively thick
oil and fairly good surface finish. Most helicopter or
final stage turboprop gears are in regime II.
In Regime Ill the EHD oil film is thick enough to
essentially avoid metal-to-metal contact. Even the
asperities generally miss each other. The high
speed gear is generally in Regime Ill. In the
aerospace gearing field, turboprop drives are high
speed and in Regime Ill. Helicopter gears are in the
high speed gear region at the input sections of the
Definition of endurance limits and regime of lubrication are outlined in clauses 5 and 8.
4.4.4 Considerations

for quality levels

Quality levels of aircraft and aerospace gears,

bearings and seals are usually as high as system
cost limitations permit or as good as can be
obtained by using todays manufacturing methods.
Aircraft engine gears are generally ground to obtain
quality 12 or better, honed to obtain good surface

High speed aircraft seals are in general carbon face

and rotating. Their designs are selected to be flat
within two Helium light bands, where each band
step measures 11.6 pin (294 pmm). In lower speed
applications, lip seals are often used.
4.4.5 Lube system considerations
Details of lube systems are discussed in clause 5.
Aircraft or aerospace gearbox components rely on
direct and pressurized lubrication for the formation
of EHD films and cooling.
Lube system design includes internal coring or
external piping, jets, spray bars, and into mesh or
out of mesh lubrication. Lube pumps,deaeration,
and filtering requirements are also considered an
integral part of the lube and cooling systems.
4.4.6 Tradeoff considerations
Completion of final design can also include a
comparative study for advanced materials vs.
conventional materials. This study includes all
rotating components and housings. Life, weight,
cost and maintainability can be compared.
4.4.7 Test considerations
Completion of any aircraft or aerospace gear
system design also includes modification of test
tools and test setups to run the following:
- Manufacturing Tests;
- Component Tests;
- Loss of Oil Tests;
- Power Plant Tests;
- Overload Tests:
- Ground Tests;
- Flight Tests.
These tests are conducted at specified environmental conditions outlined in clause 6.
Vibrations, fire resistance, weapons effects, emissions, and attitude are also integral parts of the
above defined tests.

* Numbers in brackets[ ] refer to references listed in Annex C.


AGMA 91%A94

5 Lubrication
5.1 Cooling vs. lubrication requirements
Proper lubrication of gears consists of:
a) selecting the correct lubricant;
b) ensuring that the lubricant gets into the gear
c) providing adequate lubricant flow so that heat
generated in the mesh is removed.
There are a number of other considerations in the
design of an aerospace gearbox lubrication system
but all are related to these three basic requirements.
Failure modes that can occur due to inadequate lubrication include: scuffing, micropitting and spalling.
5.1 .l Elastohydrodynamic
and lambda ratio

(EHD) lubrication

The thickness of the protective EHD oil film can be

calculated using the techniques described in appendix A of ANSVAGMA ZOOl-B88. The ratio of
film thickness to composite surface roughness is
called the lambda ratio. At a lambda ratio of one,
there is theoretically no metal to metal contact. As
the lambda ratio decreases, more and more contact
occurs. However, carburiied aerospace gears
operate successfully at lambda ratios as low as 0.4
without incurring suface damage. Aerospace gears
can operate successfully at lambda ratios below 0.4
if adequate boundary lubrication is available.
Boundary lubrication utilizes the chemistry of the
tooth surfaces, the lubricant and its additives to
provide a protective film. Since this type of lubrication is not well understood today, the designer must
match the application to past successful1 designs
operating under similar condiiions.
5.1.2 Cooling the gear mesh
In oil lubricated systems, the amount of lubricant
supplied to the gear mesh depends on the heat
generation rate. The amount of oil required in the
formation of an oil film is miniscule compared to that
required for cooling. Most aerospace lubrication
systems are designed to handle the highest heat
load and have excess capacity at all other operating
conditions. Heat generation in gears and bearings
can be estimated by various techniques [l] thru [7].

Typically, convection and radiation are ignored such

that the entire heat load is to be transferred to the
cooling oil by conduction and then removed from the
system with a separate oil cooler. When using
grease lubrication, solid lubrication and low flow
splash lubrication, heat must be removed entirely by
conduction through the housing walls or through
shafting. Cften cooling is a major limitation of these
systems. Knowing the heat load, the lubricant
characteristics and the allowable temperature rise,
the required oil flow rate can be calculated:

= M ch (T,,,,,- Td

. ..(8)


is heat generated at design point, Btu/min


is lubricant flow rate, Ib/min (kg/min);


is lubricant specific heat at ( Tout+ Th) /2,

Btu/lbmF (kJ/kgK);

Tau is average oil out temperature, F (C);


is average oil in temperature, F (C).

5.2 Lubricants
52.1 Liquid lubricants
Liquid lubrication predominates in the aerospace industry today. Many gear systems must be designed
to utilize lubricants that were originally formulated
for high temperature turbine engine applications
(MIL-L-23699 and MIL-L-7808). In some cases
the engine and gearbox use a common lubrication
system and thus must utilize engine oil. In other
cases a common lubricant has been required to prevent mixing of two diierent types of oil. These lubricants were formulated to meet criteria such as cold
flow/cold start requirements, high temperature limitations, material compatability requirements and
cost. These properties are derived from fluid base
stocks that are not necessarily ideal for lubricated
contacts in a gear drive system. Recently a new
version of these engine lubricants has been put in
service for helicopter applications (DOD-L-85734).
This lubricant isvery similarto MIL-L-23699 butaddiiives beneficial to the transmission are included.
Tables 2 through 5 list pertinent properties of the
most commonly used aircraft lubricants today.


AGMA 91%A94

Table 2 - Aerospace lubricant viscosities



400 204
350 177
320 160
212 100
210 98.9
104 40
100 37.8
-40 -40


MI L-L-23699


25.00 min*


5.00 to 5.50


25.00 min


Dexron II3


dotesReference - AFAPL-TR-71-35;
! QOD-L-85734(AS) specification
) General Motors Dexron II Specification
from MIL-L-23699D or MIL-L-7808J specifications

Table 3 - Aerospace lubricant densities










, s/ml


Jotes Reference - AFAPL-TR-71-35

1Exxon Datasheet - ET0 25
I General Motors Dexron II Specification

Table 4 - Aerospace lubricant pressure-viscosity



Reference AFAPL-TR-75-26


Dexron II3


Pressure-viscosity coefficient,
0.498 (72.2)
0.428 (62.1)
0.532 (77.2)
0.462 (67.0)
0.486 (70.5)
0.681 (98.8)
0.918 (133.2)

AGMA 911-A94

5.2.2 Greases
Greases are commonly used to lubricate actuator
gearing and gearbox components such as bearings
and splines. Several grease lubricated helicopter
transmissions are in production but are not common. The most common greasesand their uses are
listed in table 6.

Table 5 - Aerospace lubricant

specific heat values

Specific heat,
btulb OF[kJ/(kg OK)]
MIL-L-23699 or

5.2.3 Dry lubricants

Dry lubricants are widely used in spacecraft systems where liquids or greases cannot be used due
to out-gassing problems (see clause 13) and also in
aircraft systems where liquids or greases cannot be
contained. These lubricants do not provide the
same level of protection as liquids or greases.
Thus, the applied loads and sliding velocities must
be significantly lower in these systems. Table 7 lists
common dry lubricants in aircraft use today.





Reference AFAPLR-T&75-26

Table 6 - Aerospace greases*

ML Specification




Oil resistant grease

Molybdenum disulfide
Gear and actuator grease
Aircraft bearing grease
High speed bearing grease
Oil resistant grease
Aircraft grease
General purpose grease
High load capacity grease
Extreme pressure grease
Helicopter transmission grease

Tapered plug valves, gaskets

Splines, sliding steel surfaces
Bearings, gears, etc.
Ball and roller bearings to 200 000 DN
Ball and roller bearings to 400 000 DN
Tapered plug valves and gaskets
Driveshaft couplings
Bearings and gearboxes
Splines and bearings
Gearboxes, actuators
Tail rotor and intermediate gearboxes

Military Handbook, Guide for Selection of Lubricant and Compounds for Use in Flight Vehicles and
Components, MIL-HDBK-2754, May, 1969
NOTE -The above greases are not to be used in vacuum applications(see clause 13).

Table 7 - Aircraft dry lubricants*

MIL Specification




Molybdenum disulfide
Corrosion inhibiting
Air drying solid film

Dry lubricant or mix with oil

Threads, gears
Gears, flap hinges
Steel, titanium, aluminum

Military Handbook, Guide for Selection of Lubricant and Con-pounds for Use in Fli&ht Vehicles and
Components, MIL-HDBK-275A, May, 1969


AGMA 911-A94

5.3 Distribution systems

in this subclause, guidelines are described for
selection of various gearbox lubrication schemes.
The development of any new gearbox must,
however, include testing to determine the success
of the lubrication system design. There are many
factors that can unexpectedly affect system
performance. A successful lubrication system
design prevents gear tooth scuffing, meets oil
temperature rise expectations and limits power
consumption due to oil distribution and removal.
53.1 Pressure-fed
A pressure fed system, for this discussion, consists
of an oil supply line directed at a gear mesh which is
fitted with a metering orifice.
5.3.1 .l Oil pressure requirements
Oil pressure is required to propel lubricant into a
rapidly moving ,gear mesh. An oriiice is used to
convert oil pressure into kinetic energy creating the
oil velocity required to penetrate the mesh. By
properly selecting both the oil pressure and the
orifice (jet) diameter, the flow rate and oil
penetration depth can be controlled.
53.12 Analysis for pitch line velocities to approximately 25 000 feet per minute (127 mkec)
For gear meshes operating at 25 000 ft/min or less
the technique described in [8] can be used. Oil
penetration depth into the mesh is simply a time of
flight calculation. At these speeds the jet still acts as
a solid stream and is not significantly affected by air
turbulence. Ati long as the oil flow is not interrupted
by the passage of another tooth, oil will penetrate to
the depth as determined by this method. Analysis for pitch line velocities greater
than 25 000 feet per minute (127 m&c)
Standard analytical methods have not been confirmed above 25 000 feet per minute. Windage of rotating parts must be considered to ensure proper
targeting of the oil jet. Development testing of the
unit is required to determine the adequacy of the lubrication scheme employed. Oil jet orientation
Oil isfrequently supplied toward thedisengagement
side of the mesh on the driving side of the tooth.
Power losses are minimized, cooling oil will reach

cooling is maximized. In some cases only a small

amount of oil is required to provide adequate
into mesh lubrication can increase
churning losses due to the presence of excess oil in
the mesh.
In high power applications it may be necessary to
use multiple oil jets to provide the required oil flow
and still maintain jet velocity. Oil jets can be placed
on both the out of mesh and into mesh sides to obtain the required oil flow rates. In this case the intomesh jet should supply a lesser flow.
5.3.2 Splash
Splash lubrication is used when heat generation
rates are low compared to cooling available either
through conduction through the gear or through
convection to a nearby lubricant. A first order heat
transfer analysis can be performed to determine the
cooling available from the gear and the lubricant.
Splash lubrication can be provided by dipping the
gear through a pool of oil or by placing the gear in
the path of a stream of oil exiting from another mesh.
Development testing is typically performed to determine adequacy of the lubrication scheme.
Consideration should be given to pre-lubing gearboxes of this type to ensure that oil is available during startup.
5.3.3 Mist
Mist lubrication is similar to splash lubrication except that the cooling provided by the oil is very low.
The oil available in the mist is sufficient only to allow
generation of an oil film. In some cases an air/oil
mist is created with the intent that the air will provide
cooling of the mesh.
5.3.4 Grease
Grease is used for gearbox lubrication when a
sealed area can be provided to prevent contamination and lubricant loss. The advantages include low
cost and simplicity particularly when it would be
diicult to introduce a liquid lubricant system. Heat
generation in these components must be low
compared to the heat sink available through
conduction to the housing walls and the shafting to
prevent high operating temperatures.
Sealed, grease lubricated gearboxes have been

the tooth flank at its highest temperature thus

used in helicopter tail rotor and intermediate gear-

promoting heat transfer and the length of time for

box applications.


In addition to the advantages

AGMA 911-A94

stated above for bearings and spiines they are also

less vulnerable in combat situations.
Flap actuator gear drives are another common
aerospace application for grease lubrication.

transmission and also has less tendency to chum

the oil. Wet sumps may not be practical for some
applications due to the increased frontal area
required or other envelope limitations.

5.3.5 Powder lubrication

5.4.3 Oil deaeration

Powder lubrication is being investigated for very

high temperature applications where only dry lubricants can survive. in these systems dry lubricants
are blown into bearings and gears as an aerosol to
lubricate and cool the system.

Oil foaming is a major concern in aircraft gear

systems. Air is easily trapped in oil due to the mixing
action that occurs during high speed rotation of
gears and bearings. Foaming can be controlled by
allowing air to escape naturally in an oil tank,
providing deaeration trays in the oil tankor by using
air/oil separators. Oil circulation rates must be
selected to provide time for air to escape in an oil
tank or tray. Since this usually requires a larger oil
tank it is sometimes necessary to supplement the
natural deaeration with an air/oil separator.

5.4 Lubrication system design considerations

5.4.1 Common vs. separate lubrication systems
In some cases the designer may have the choice
between a self contained lubrication system or an
external system. Frequently, however, the gearbox
lubrication system must becommon with the engine
or other equipment and the resulting complexities
must be considered. Table 8 lists some of the advantages and disadvantages of each type of system.
5.4.2 Dry sump vs. wet sump
in a dry sump system, oil is stored in a separate oil
tank, not part of the transmission housing. In operation, oil supplied to transmission components drains
to scavenge ports where the air/oil mixture is
pumped out to the tank.
The wet sump system integrates the oil tank into the
transmission housing, typically at the bottom.
Scavenge pumps are required only for areas that
are diiicult to drain. Otherwise, gravity is used to
return oil to the tank area. Oil pump inlets are
designed to remain covered with oil during all
attitudes of operaion.
The wet sump system offers the advantages of no
external plumbing for connecting the oil tank to the

5.4.4 Oii scavenging systems and oil baffles

In order to prevent heat generation due to churning
of excess oil near rotating bearings, gears or seals,
it is frequently necessary to use oil scavenge pumps
to return oil to the sump or tank. If this excess oil is
not removed several problems can arise: excessive
power consumption due to oil churning, high oil
temperatures, oil foaming, oil leaks through seals,
and larger oil tank capacity requirement due to
change in level on start-up.
In the design of the scavenge system the designer
must consider attitude and altitude requirements of
the system. The scavenging action must continue
through all aircraft maneuvers and thus through a
range of system attitude alignments. This requirement frequently results in multiple oil scavenge
pumps with oil pick-up passages placed in strategic
locations. It is generally not feasible to use one
pump with multiple scavenge pick-up points since
only air will be removed if any one passage
becomes exposed to air.

Table 8 - Advantages and disadvantages of a common engine and transmission

lubrication system
Common system

Separate systems

Lower weight - common parts

Engine oil must be used
Transmission failure can affect engine

Increased weight - redundant parts

Transmission oil can be used
Transmission failure is self contained

External plumbing required

Self containedplumbing


AGMA 911-A94

Scavenge oil pumps are typically rated at two to

three (and sometimes more) times the supply pump
flow rate. This ensures that oil will not build up in
critical areas under adverse operating conditions.
One of these conditions occurs at altitudes where
air is less dense and is easily dissolved into the oil.
Removal of this foamy mtiure requires additional
flow capacity since the volume of this mixture is
greater than solid oil.
Design and placement of oil baffles can also affect
oil scavenging and gearbox performance. The
purpose of an oil baffle is to prevent oil from
becoming entrained in rotating gears and bearings
and to remove cooling oil that has done its job.
Many of the same problems discussed above will
occur if oil is allowed to flow into rotating parts.
An oil baffle diverts oil toward the sump or scavenge
port to prevent a build-up of oil within the housing. A
baffle can also be used to control air windage and its
effect on oil foaming. A baffle can be cast into the
housing or fabricated separately and bolted into
position. The baffle should be tested to determine
its natural frequency to ensure that it will not be
subject to high cycle fatigue in the high vibration
environment inside the gearbox.
5.4.5 Pressure drop in oil passages
Consideration should be given to pressure drop in
oil passages, particularly if the length of the
passage is long, if there are a number of sharp turns
or if the surface roughness of the passage is poor as
in cast passages. If this is not done, the calculated
flow rate could be significantly less in operation. A
rule of thumb used in the past for sizing passages
has been to design for a velocity not to exceed 15
ft/sec (4.6 rn/sec) on the pressure side and 5 ft/sec
(1.5 m/se@ on the suction side.
5.4.6 Oil sampling
If oil sampling capability is required it must be
designed into the system at an early stage.
Frequently this feature is added to an existing
system and good results are not possible. In order
to obtain a representative oil sample, the oil tap
should be located at the end of a line just before a
turn is made. When the sample is taken, oil from the
main stream is obtained. In many cases an oil tap is
placed in a region where debris is allowed to build
up and the sample is artificially contaminated.


When taking an oil sample it is good practice to

clean the tap priorto taking the sample and to let the
oil flow for several minutes before taking the actual
5.4.7 Fill/drain considerations
It is often possible to introduce more contaminants
intoagearsystem during top-off than during normal
system operation. This is particularly true for
systems with fine filtration. A fine filtration system
cleanses the entire system over a period of time to
very low contaminant levels. Opening the system
during top-off may allow debris to fall into the
system. Control of the top-off function can partially
be aided by design and placement of the fill
mechanism but is primarily controlled by training
maintenance personnel.
Oil changes are common in systems with coarse
filtration. Change intervals are determined from
past experience with similar systems. Oil changes
in systems with fine filtration are less frequent. The
cleanliness levels found in fine filtration systems are
maintained continuously. The lack of circulating
debris tends to reduce the amount of new debris
that is formed. In systems in which the engine and
gearbox are common, oil consumption by the
engine requires a routine tovff
of new oil. This
replenishes any additives that may have been lost.
5.4.8 Vulnerability/safety
Some gearbox specifications include the requirement for an emergency lubrication system
independent of the primary system.
emergency system allows safe operation of the
aircraft for a limited time in the event of the loss of
the primary system. The degree of sophistication in
these systems can range from a complete
secondary pumping system to a very simple gravity,
In addition, some military
oil drip system.
transmissionsare required to operate for a specified
period of time after loss of the lubrication system.
5.4.9 Low temperature operation
Most aircraft gearboxes must be designed for
operation at a very low temperature, typically4
(-54 C). Since lubricant viscosities are quite high
at this temperature, it is diiicult to circulate oil in the
lubrication system. The requirement led to the
development of the MIL-L-7808 lubricant with its
lower cold temperature viscosity. Even with this

AGMA 911-A94

lubricant, oil heaters may be required to aid starting.

It is difficult to pump lubricant during start-up at low

points such as a main distribution line leading to

gear mesh oil jets. Screens are also used to protect
scavenge pumps.

5.4.10 Oil coolers

5.5.3 Filter rating nomenclature

Some type of oil cooling system is required for most

aerospace gear systems. Convective heat loss
through the housing may not remove enough heat
under all flight conditions to maintain acceptable oil
temperatures. Typical oil cooling systems include:

A typical filter rating might be: 10 micron absolute,

Beta=1 00. This terminology means:

- ram air/oil heat exchangers;

- air/oil heat exchanger with blowers;
- fueVoil heat exchangers for lower heat load
Oil temperature is sometimes controlled in these
systems by varying the amount of air or fuel
supplied to the coolers. A typical aerospace lube
system schematic is shown in figure 4. Some
gearboxes that have low power consumption can be
operated without oil coolers if the heat loss is
sufficient to keep operating temperatures within
acceptable ranges.
5.5 Filtration
5.5.1 Filter types
There are three types of aircraft filters in use today
characterized by the media used to trap debris:
steel mesh, paper and fiber mesh. Steel mesh filters are being utilized less due to:
-the maintenance required to clean them;
-the loss in effectiveness after cleaning;
- cost.

Paper media filters are generally used for coarse

filtration levels. Throw-away fibrous media filters
are becoming more common as fine filtration is
becoming more popular.
5.5.2 Screens
Screens or last-chance-filters are used to protect
system components from large pieces of debris,
The mesh is usually coarse and is intended to stop
debris from a failure that occurs upstream of the
screen but downstream of the main filter. Screens
are typically placed ahead of critical lubrication

The fitter will reduce the number of particles

greater than or equal to 10 microns in size
entering the filter by a factor of 100. This equates
to an efficiency of 99 percent. Or, 1 percent of the
particles greater than 10 microns will pass
through the filter.
Nominal filter ratings are also used today but are
less stringent. Nominal efficiencies are typically 90
to 98 percent.
5.5.4 Fine filtration
Pine fittration is becoming more common today for
aircraft engine and gearbox systems for several
reasons: longer bearing and gear life due to lower
levels of debris, fewer or no oil changes and potentially longer filter life. Fine filtration requires a 3 to 5
micron absolute rating with a Beta factor of at least
100. Fitter life has been found to increase after the
initial system clean-up phase. Once the system is
clean, it tends to stay clean.
5.5.5 Filter location
Filters can be located on the pressure side of the
system, the scavenge side or both. Selection of the
site depends on the designers philosophy, e.g. a
pressure side filter will always protect the component, whereas a scavenge filter will protect the heat
exchanger, sump or tank, and will not proliferate the
debris. The overall system design must be considered to reach the best decision on filter location.
5.6 Oil pumps
The three most common aerospace gearbox oil
pumps are the gear pump, the vane pump and the
gerotor pump . All are positive displacement pumps
since output flow is a function of fifed internal
geometry. Flow increases linearly with shaft speed.
However, the vane pump can be designed to
provide a variable flow rate.

AGMA 91%A94

Oil tank

From nacelle

Prjp gearbox lube and scavenge pu$ps

To dacelle

Gear mesh

Gear mesh


kf Gear mesh


)-I Bearings



Prop gearbox sump


Accessory gearbox sump


Screen mesh


Regulating valve


By-pass valve

Figure 4 - Typical aerospace lubrication system schematic

56.1 Gear pumps
A gear pump moves a fixed volume of oil from the
inlet side to the outlet as shown in figure 5. Plow
volume is a function of tooth depth and width. As
resistance to flow increases at the outlet, the power

gear versions of this type of pump can increase the

flow with little increase in pump size.

5.62 Vane pumps

Vane pumps consist of a rotor mounted off center in
a circular cavity (see figure 6). Vanes in the rotor are

required to rotate the input shaft increases and

forced out due to centrifugal force under rotation

discharge pressure increases accordingly. Three

and contact the cavity walls.


The amount of

AGMA 916A94

eccentricity determines the volume of flow per

revolution. Variable flow rates are achieved by
allowing adjustment of the eccentricity during operation. Variable displacement pumps are more
efficient since excess flow (and thus excess power
loss) is eliminated.

Figure 5 - Spur gear pump

5.6.3 Gerotor pumps

Housing body/cam ring



Figure 6 - Vane pump

Gerotor pump action is shown in figure 7. Oil is

drawn into an area equivalent to one tooth space at
position 1 and is rotated and then forced out due to
compression as it moves toward the exit at position
4. Both rotors rotate but the inner rotor driies the
Spring type pressure regulators are often used to
maintain a constant oil pressure under varying conditions of operation, e.g. speed and temperature.
These regulators are commonly built into the pump
body. Oil flow rates are then controlled by selecting
orifice (jet) sizes to supply the correct oil flow at the
regulated oil pressure.

5.7 Lube system condition monitoring

5.7.1 Chip detectors
Magnetic chip detectors are used to collect debris
circulating in the lubrication system. In some
systems the pilot receives an indication when the
number of particles collected is sufficient to trigger a
signal in the detection circuit. In other systems the
chip detector is visually checked as a maintenance
function. Chip detector operation is very sensitive to
systemconfiguration and sensor location. Testing of
chip migration in the housing is recommended for
placement of chip detectors.

Figure 7 - Gerotor Pump


AGMA 91%A94

There are many types of chip detectors but their

function is similar to that just described. The burnoff chip detector was developed to reduce the incidence of nuisance indications during flight. This device automatically passes a strong current through
the collected chips, melts them, and may provide an
indication to the pilot. This will clear the indicator. If
a true failure has occurred chips will collect again
causing another indication shortly thereafter.
Another type of chip detection system under development has the ability to count particles of several
sizes and track the build-up of debris. If the rate of
build-up is rapid, there is a good chance that a failure has occurred.
5.7.2 Temperature/Pressure sensors
Lubricant temperature and pressure are the primary
parameters used to track system condition. These
sensors must be placed in the system correctly to
obtain good results. As an example, a temperature
sensor that is to detect the temperature of the oil exiting the gearbox must be placed in an area where
good lubricant flow is taking place. Otherwise, it
may just indicate air temperature.
5.7.3 Spectrometric oil programs
Spectrometric oil analysis programs have been
used for many years to detect wear of components
in aircraft systems. The procedure consists of taking oil samples at regular intervals and analyzing
them for metal content in a laboratory. These programs have shown varying degrees of success.
Due to the size of the particles analyzed, this
method may not work in fine-filtration systems.
6 Environmental


This clause illuminates the environmental issues

likely to be encountered by an aerospace gearbox
during its lifetime, including storage, transport, and
Data on the occurrence of environmental extremes
may be found in several publications. MILSTD-210, Climatic information to Determine Design and Test Requirements for Military Systems
and Equipment provides data on free air (i.e. ambient) conditions. During the development phase of a
gearbox, MIL-STD-810, Environmental Test Methods and Engineering Guidelines, can be used to

MlL-STD-462, Measurement of Electromagnetic

Interference Characteristics.
Another source of environmental information is the
NASA Technical Memorandum 82473, Terrestrial
Environment (Climatic) Criteria Guidelines for Use
in Aerospace Vehicle Development.
6.1 Ambient temperature effects
The operating ambient temperature range for the
gearbox will be determined by the aerospace system application operating envelope in conjunction
with data presented in the International Civil Aeronautical Organization (ICAO) Standard Atmosphere. [Note that the ICAO Standard Atmosphere is
identical to the US Standard Atmosphere for altitudes below 65 617 feet (20 000 meters) .]
The gearbox should be capable of operation at
continuous rated power or less at any ambient
temperature within the specified operating envelope. It is important to note that at temperature
extremes material characteristics and properties
can be significantly affected. For example, materials with dissimilar coefficients of thermal expansion
(e.g., aluminum versus steel) will significantly
influence operating fiis and clearances within the
gearbox and need to be adequately considered in
the design. Also, at very low temperatures lubricant
viscosity will be orders of magnitude higher than at
typical operating temperature. As a result, lubricant
flow through heat exchangers may not occur even
after the remainder of the gearbox is at normal
operating temperatures. The net consequence is
that power loss through the gearbox will be significantly increased.
In addition, increased brittleness and notch
sensitivii can become an issue at very low
temperatures. For example, the fracture toughness
of gear materials can be reduced by more than
twenty fiie percent when going from room
temperature to -65F (-54%). Other potential low
temperature problems are loss of O-ring resilience,
resulting in leakage of lubricant from the gearbox
and contraction of the gearbox at a greater rate than
the gears and bearings, causing abnormal loading
ofthesecomponents. Refertoclause13forvacuum
environment applications
At the other extreme, lubricant viscosity will be very
low leading to the possibilii of very thin oil films in
the gear meshes and in the internal bearing con-

gain additionalinsights. Electromagnetic


tacts. If the oil film thicknessis too small,allowable

ence test and analysis requirements are covered in

compressive stress limits will be reduced for the


AGMA 911-A94

gears and bearings and the probabilii of gear tooth

scuffing (scoring) or bearing surface distress will be
significantly increased. The judicious selection of
oils with appropriate properties at both extremes
(see clause 5 for detailed information) can help
avoid these potential problems as can the proper
design and sizing of oil coolers to maintain a desired
oil operating temperature.
Another consideration associated with very high
ambient temperature operation is the retention of
adequate surface hardness for both gears and
bearings. Sufficient loss of hardness (e.g., two
points below design minimum values) can radically
degrade compressive fatigue life. Therefore,
retention of adequate hardness is a very important
consideration in the proper selection of gear and
bearing material.

6.3 Attitude effects

The gearbox operating attitude limits will vary
greatly depending upon the particular aerospace
application. For reference, figure 8 shows typical
limits for military aircraft. The aircraft/gearbox
should be capable of continuous operation in the
unshaded area and capable of some period of
transient operation in the shaded area. Sometimes
diierent duration requirements are specified depending upon whether the system is at zero g
(acceleration due to gravity) or negative g operating
conditions. At any attitude, the gearbox should be
capable of operation at continuous rated power or
less. In addition to operating requirements, the
gearbox should be capable of being stowed or
transported across the specified attitude spectrum.
* Above

Special mention is madeforaerospace applications

required to operate at cryogenic temperature (i.e.,
at liquified gaseous conditions). At these extremely
low temperatures, very specialized design practices, materials, and lubricants are needed to ensure proper gearbox operation.
6.2 Ambient pressure effects
The operating ambient pressure range for the gearbox will be determined by the operating envelope of
the particular aerospace application. Typical factors affecting pressure include operating altitude of
the aircraft, rates of climb/dive, and the effects of nuclear or conventional weapons. At extremes, pressure changes, at a rate of 1000 psilsec (6.895 x 1O6
Pa/se@ with magnitudes of 2.5 psi (1.724 x 1O4Pa)
and 70 psilsec (4.826 x lo5 Pa/set) with magnitudes of 5 psi (3.447 x 1O4Pa) can be anticipated.
The gearbox should be capable of operation at continuous rated power or less at any ambient pressure
or pressure change within the specified operating
envelope. Design of the gearbox should include
consideration of the possible pressure changes on
performance. Pressure differentials across gearbox
seals can fluctuate and even reverse in flight if not
properly designed for and cause seal contamination
and oil leakage. Likewise, venting of the lube system is often accomplished within the gearbox
through breather pressurizing valves which need to
be designed for all possible pressure variations.
Adverse consequences include loss of oil, oil entrapment, and oil foaming, all of which can lead to
early failure of the gearbox components.


,Roll &-rgle

A *Below A

* Reference to ground
n - Test points
Notes 1. The gearbox shall be capable of operating at all possible conditions; however, for the purpose of defining
the direction of acceleration vector from the engine CG,
the figure assumes no acceleration other than grav.w.
2. Gearbox centerline perpendicular to plane of paper.
3. Continuous operation in clear area.
4. Thirty second operation in shaded area.

Figure 8 - Typical gearbox attitude limits

The requirements for variable attitude limits derive

from the fact that the aircraft application is required


AGMA 911-A94

to climband bank, and formilitaryaircraft, tosustain

air combat maneuvers and terrain following operation and possibly undergo inverted flight. If the oil
system for the gearbox is not properly designed to
operate within the specified attitude limits, serious
consequences for the gearbox can result. For example, extended operation in zero or negative g
conditions can result in oil system malfunctions
which can effect the gearbox such as oil starvation,
gearbox flooding, oil foaming problems, and seal
leaks. Internal pressurization techniques are often
used as solutions to these problems. Special provisions, such as an auxiliary lube system, may be
necessary in the gearbox lubrication system for extended inverted flight operation. Placement of suction/feed points to the pumps can alter the attitude
envelope. See 5.4.4.
6.4 Contaminant effects (water, corrosives, dirt,
dust, and sand)
Due to worldwide all weather operation requirements for many aerospace systems, gearbox operation in the presence of contamination has become a necessity. For example, aircraft operating
out of airports near arid parts of the world often takeoff and land in conditions where the air contains
sand, dust, and dirt. Water gets ingested during
rainstorms. The fact that the lubrication system
which services the gearbox has to breathe means
that atmospheric water and vapor is ingested into
the gearbox. This moisture may or may not contain
salt which presents a whole host of corrosion related problems. The incidence of contamination is
so prevalent that specifications have been developed for them. The table below shows the data for
US Specification Sand. The specified sand should
include 90 percent silicon dioxide (SiO;z)of angular
structure with total particle size distribution by
weight as shown in table 9.
Table 9 - Particle size distribution by weight
Size, fin

Size, CPM

o-2 953
2 953-4 921
4 921-7 874
7 874-15 748
35 433-39 370

(75-l 25)
(900-l 000)

by weight
3.5 lk 0.5

Using the specification contaminant at some predetermined concentration level, qualification tests are


conducted to validate the adequacy of the gearbox

or other component design to operate for some
extended period of time. For the required duration
the gearbox should be capable of operation without
any degradation in performance. Obviously, the
best wayto handle the contaminant is to preclude its
introduction into the gearbox by prudent design. For
the larger contaminants, like sand, this may be
possible. However, for water or dirt this is probably
not possible and other solutions have to be found to
accommodate them such as filters to remove the
dirt after it has entered or protective coatings to
resist the deleterious effects of water.
6.5 Vibration/Shock effects
This subclause covers vibration and shock effects
on the gearbox ensuing primarily from external
sources. Those initiated internally within the gearbox are covered extensively in clause 7. The random vibration environment which an aerospace
system gearbox might be subject to can originate
from the following sources:
- Turbulent aerodynamic airflow along external
surfaces of the airframe structure;
- Engine noise impinging on the airframe structure;
- Rotor/propeller blade induced effects;
- Airframe structural motions causing
sympathetic gearbox response;
- General aircraft motions caused by such
factors as runway roughness, landing, and
- General motions encountered during
transport of the gearbox, such as by rail, ship,
or truck;
- For military systems, the gun blast pressure
impinging on the aircraft structure from high
speed repetitive firing of installed guns and vibration from antiaircraft fire (flak).
Typical vibration spectra tested for in aircraft are
shown below in table IO and figure 9 for propeller
aircraft and turbine engine equipment, and in figure
10 and table 11 for helicopters.
The shock environment is often a consequence of
the following effects: aircraft launch/catapult and
landing; aircraft crash; handling, such as dropping
of the gearbox; and transport of the gearbox.
Typical shock pulse levels and durations tested for
are shown in figure 11. Generally, specific requirements are specified by the customer. These may
diier from those in the figure.

AGMA 911-A94

Table 10 - Suggested functional test levels for propeller aircraft and turbine eng,ine equipment
Equipment location*- 3

Vibration level (Ll) at F1 4$5, g*/Hz


In fuselage or wing forward of propeller

In fuselage or wing aft of propeller
In engine compartment or pylons
Equipment mounted directly on aircraft engines

1 Fl= fundamental excitation frequency; Fi= source frequency (i = 1 -4), FZ = 2 Fl, F3= 3 FI, F4= 4 F1
2 When panels and racks are not available for equipment installed on vibration isolated panels or racks, or when
the equipment is tested with isolators removed, use fuselage or wing forward of propeller category with levels
reduced 4 dB.
3 Increase test levels 6 dB for equipment mounted on fuselage or wing skin within one propeller blade radius of
the plane of the propeller disc. For all other skin mounted equipment, increase levels by 3 dB.
4 Bandwidth vibration around each Fiwill equal f 5% F fpr constant-speed excitation. When excitation is not
constant-speed, bandwidth will encompass operating speeds for cruise and high power operation.
5 Fl= 68 Hz for most C-l 30 aircraft.



-6 dB/octave
octave = 2 to 1
frequency range

6 -01 n




l .










_---_-_ __
8 I





Frequency (Hz)

Frequency (Hz)

B. Turbine engine equipment spectrum

A. Propeller aircraft spectrum

Figure 9 - Suggested vibration spectra for propeller aircraft and turbine engine equipment


*; $02


Note: See table 11 for

Lt , Lf2, L13and L/4 g-levels.

L3 L4











instrument panel
External stores
On/near drive
train elements

w, , g*A-lz

Ft , Hz



Frequency (Hz)

Figure 10 - Suggested vibration spectrum for equipment installed on helicopters


AGMA 91%A94

Table 11 - Suggested functional test peak levels for equipment installed on helicopters
Equipment location

Source frequency (5) range, Hz Peak vibration level at F,, L,, gs



0.1 F,
6.5 - 0.1 F,


0.07 F,
4.55 - 0.07 F,

External stores


0.15 Fx
9.75 - 0.15 F,

On/near drive train elements*


0.1 F,
5 + 0.01 F,



Source frequency of interest = Fr, Fe, FJ, or F.4
Fj = Fundamental source frequency
Fp = 2F1 ; F3=3Fj
; F4 = 4F1
Upon determining values of Fl, F2, F3, or F4 (figure lo), select the appropriate source frequency range for each
when determining peak vibration levels. The source frequency ranges are not presented in order of FI = F4.
Ft, F2, F3 and F4 must be determined from drive train areas for the particular helicopter. Note (1) is then


Ideal sawtooth pulse

Tolerance limits

Operation test

Crash safety

Minimum peak value, (P) gs


Flight vehicle



Note - The oscillogram shall include a time

about 30 long with a pulse located approximately in the center. The peak acceleration
magnitude of the sawtooth pulse is P and its
duration is D. The measured acceleration
ulse shall be contained between the broken
line boundaries and the measured velocity
change (which may be obtained by integration of the acceleration pulse) shall be within
the limits of K fO.l I$ , where X is the velocity change associated with the ideal pulse
which equals 0.5DR The integration to determine velocity change shall extend from 0.4 D
before the pulse to 0.1 D after the pulse.

Nominal duration, (0) ms

Flight vehicle



1 Shock parameters recommended for equipment not shock-mounted and weighing less
than 300 pounds (136 kg).
2 Equipment mounted only in trucks and semitrailers may use a 20 g peak value.

Figure 11 -Terminal-peak sawtooth shock pulse configuration and its tolerance limits
(for use when shock response spectrum analysis capability is not available)



Some potential adverse effects of shock and

vibration on a gearbox include cracking of gearbox
housings and covers, breaking of gearbox mounts,
cracking of oil supply and scavenge lines, misalignment of internal components, and brinelling/
false brinelling of the gearbox bearings and gears.
Several means are typically employed to minimize
these adverse effects. For example, vibration
damping coatings can be applied to gearbox
housings, elastomeric dampers are used for gearbox mounting, pneumatic or hydraulic mounts are
employed during transport, and oil lines are amply
supported to preclude the presence of potentially
harmful resonant vibration frequencies.
Qualification tests should be conducted on the
gearbox at the appropriate vibration and shock
levels and durations (again depending on the
particular aircraft application) to validate the gearbox design. Because of potential airframe effects,
testing of the total aerospace system is necessary
for final evaluation.
6.6 Fire resistance requirements
The capability of the aircraft and its many
components including the gearbox to have
adequate fire resistance is critical to overall
aerospace system operational safety. It is essential
that any system component containing flammable
fluids, such as the gearbox with its oil lubrication
system and fuel-oil heat exchangers, will not
contribute to an aircraft fire and will allow sufficient
time to isolate the fire zone. Gearbox seals, covers,
and ftiings are primary areas of concern for leakage
and potential fire initiation sources.

resistance requirements. These tests are conducted while conveying the oil within the gearbox at
the lowest oil flow rate, highest ambient system
pressure, and the highest oil temperature expected
over the complete gearbox operating range.
6.7 Electromagnetic effects
Electromagnetic compatibility between any gearbox electrical/electronic component and those of
the rest of the aerospace system is essential for
For example,
proper and safe operation.
electromagnetic interference could conceivably
cause a malfunction of the gearbox condition
monitoring system such as that used to monitor
gearbox vibration levels or to detect the presence of
foreign or wear-generated contamination. Additionally, other components including gears and
bearings may become magnetized due to electromagnetic effects and attract harmful ferrous-based
contamination that can lead to wear and premature
To mitigate electromagnetic effects,
shielding can be used with electrical leads and/or
with the electronic component itself. During periodic
gearbox inspections and teardowns, procedures
often include passing bearings and gears through
demagnetizing coils to preclude rebuilding the
gearbox with magnetized components. Electromagnetic interference testing methods, such as
those prescribed in MlL-STD-462 for military
aircraft, will ensure compatibility between all aircraft
electrical and electronic components.
6.8 Nuclear, biological,

and chemical (NBC)

Reduction in mechanical properties can lead to

permanent deformation of the gearbox resulting in
degraded gearbox performance or causing oil
leakage paths. Qualification tests of sufficient
duration should be conducted on the complete

Some aerospace system vehicles may have to

operate in a strategic or tactical nuclear warfare
environment. Therefore, all system equipment
including the gearbox must be optimized to withstand that environment. Potential nuclear weapon
effects include the following: blast/gust, thermal,
transient radiation, and electromagnetic pulse (see
6.7). In addition, chemical and biological weapons
are becoming more widely available and have been
used in recent warfare. Thus, the gearbox must be
capable of sustaining operation in a total NBC
environment. NBC contamination can have direct
effects upon the critical properties (e.g., physical,
chemical, mechanical, thermal, or electrical) of the
In addition, materials or
materials employed.
procedures used to decontaminate equipment
items may also degrade the properties and reduce

gearbox assembly, includingappropriatefittings, to

gearbox/drive system capabilities. To overcome

validate its compliance with specification fire

NBC effects, selection of materials known to be

In the design of a gearbox, materials and fire

resistant coatings are selected that offer adequate
resistance to fire and which can also maintain
appropriate mechanical strength properties. As
examples, aluminum and magnesium, which are
often used for gearbox housings, have repeatedly
demonstrated more than adequate fire resistance
and sufficient resistance to thermally degraded
mechanical properties.


AGMA 9ll-A94

resistant to the expected NBC condition or specific

design protection, such as nuclear hardening, must
be provided. Material resistance can be established
either through testing or by analysis of effects on
similar materials or designs. Use DOD-BTU-21 69
for reference and additional information, especially
for military applications.

7 Vibration and noise

7.1 Causes of gear vibration
Gearbox vibration is the result of a complex interaction among the gears, bearings, shafts, and housings which make up a gearbox assembly. In order to
reduce this vibration, the precise mechanisms by
which it is generated must be understood. Several
mechanisms have been identified which are primary contributors. There are also some secondary
contributors which may amplify or reduce vibration
levels. Figure12 is a schematic representation of
the way vibration and noise passes from the source
at the teeth to the foundation and the air. Note that
the path includes the gear bodies, shafts, bearings
and housing.

7.1 .I Transmission error

One of the most significant sources of gear vibration
is the result of the nonuniform transfer of torque between mating gears due to variations in the mating
parts, the elastic deformation (bending and contact)
of the gear teeth under load, and the deformation of
the tooth backup rim and web. The variations can
include pitch line runout, tooth spacing variation
(both tooth-to-tooth and accumulated pitch errors),
and profile variation. Each type of variation will produce a characteristic vibration pattern. For example, pitch line runout will produce a strong once per
revolution excitation while spacing variation will produce excitation at mesh frequency (rpm times number of teeth) and harmonics of mesh frequency. The
resulting nonuniform transfer of torque produces a
dynamic force which in turn excites torsional/lateral/
axial vibratory modes of the gear shafts. The shaft
response produces displacements at the shaft support bearing locations. These excite the housing
and cause it to vibrate at all mesh frequencies. The
dynamic characteristics of the internal components
may magnify this excitation. Furthermore, the dynamic characteristics of the housing may also amplify its displacements and resulting noise.

r noise


Airborne structure noise

Structure-borne noise

Figure 12 - noise and vibration paths


AGMA 911-A94

Due to deterioration after many hours of service,

many gearboxes will exhibit increases in noise or
vibration levels relative to the levels when they were
new. This effect is usually the result of changes
occurring in tooth profile accuracy and even in tooth
spacing accuracy. These changes are the result of
surface deterioration due to either wear or localized
pitting, and are most common with through hardened gears, but can occur even with case
carburized teeth. The condition is aggravated if the
operating oil film is marginal, either due to low
viscosity or low oil film thickness.
7.1.2 Unbalance
All rotating machinery is subject to vibratory excitation due to the dynamic unbalance of the rotating
components. Any unbalance will produce a rotating
force with a frequency equal to the rotational
frequency of the component. This force will be
proportional to the square of the rotational speed
and will cause a response from all the parts of the
gearbox which depends on their dynamic characteristics, i.e., natural frequencies and damping.
7.1.3 Misalignment with connected equipment
Many gearboxes have flexible couplings to provide
both input and output connections. Misalignment of
the connected equipment to these couplings is a
possible source of vibratory excitation. In general, if
flexible couplings are used which are not of a constant velocity type, misalignment will produce a torsional excitation with a frequency of two times the
rotational frequency. The gearbox can respond to
this excitation just as to any other vibratory source.
7.1.4 Resonance
Although resonance is not a vibratory source, since
it represents the coincidence of an excitation
frequency with a system natural frequency, it is
probably the most common cause of vibration
problems. Once the gearbox is designed, the
frequency characteristics of vibratory sources in the
system are well defined. If any of the gearbox
components has a natural frequency which is close
to one or more of these excitation frequencies or
harmonics of them, vibration or noise can occur.
One interesting example of this is the so called
beat phenomenon where a gearbox will exhibit
noise at a frequency equal to the difference
between a system characteristic frequency and a
nearly equal exciting frequency. Another example

is the case where a turboprop gearbox must avoid

natural frequencies at or near the propeller blade
passing frequency; i.e., number of blades times
propeller rpm.
7.1.5 Entrainment
Another source of vibration which is generally of sig
nificanceonly on very high speed units or those with
very wide face widths and moderate to high speeds
is that created when the mixture of air and oil which
occupies the space between adjacent teeth is
forced out by the entrance of a tooth on the mating
gear. In a high speed mesh, the speed of the air-oil
mixture as it exits the mesh area is many times the
pitchline velocity of the gear set and may, in some
cases, reach the speed of sound. At the speed of
sound, this effect dominates the overall noise level
of the box.
7.2 Consequences of vibration
72.1 Structural issues
7.2.1 .l Fatigue
One of the most severe consequences of excessive
vibration in a geared system is the possibility of
fatigue failure of internal components due to
vibration induced stresses. When a critical
component has a resonant frequency within the
operating range there is a risk that high stressescan
occur and some method of reducing the response
should be undertaken. This can include redesigning
the gear to change its natural frequencies or adding
some kind of damping. Fretting
Fretting occurs when parts are in intimate contact
and are subjected to microscopic motion. Although
fretting can occur even if vibration levels are low, the
risk of serious fretting problems is greatly increased
when the parts are also subjected to high vibration. Fasteners and lockwire
A high level of vibration can be a significant
contributor to locknut loosening, breakage of lockwire or broken tab washers in gearboxes. Brackets
If external brackets are used to provide support for
accessories or for gearbox mounting, they can be
particularly susceptible to vibration. Since brackets
are often designed with simple structural shapes
connected together, they can have high stress con-


AGMA 91%A94

centrations at the intersections of these shapes.

When subjected to vibratory loads, these intersection points can become highly stressed areas.

b) Natural frequency analysis of the gear blanks

to determine the mode shapes and frequencies
of these components; Lube system components

c) Determination of the dynamic gear loads applied to the components;

Internal lube system components such as nozzles

and windage baffles often have low natural
frequencies and as such can be susceptible to
vibration problems. Particular attention should be
paid to attachment points and any bolt flanges.
7.2.2 Noise
There is a very close relationship between mechanical vibration and audible noise. If all mechanical
vibration were eliminated, noise also would be
eliminated. However, it is not necessarily true that
reducing the vibration of any single component will
reduce noise. On the other hand, selectively
increasing the vibration of certain components, if
coupled with node relocation, can produce a noise
The noise which is ultimately radiated is composed
of two components: airborne and structure borne
noise (see figure 12). Airborne noise is transmitted
directly from the source. Structurebome noise is
due to the transmission of the vibratory energy
through the internal structure of the gearbox and
into the external supporting structure. The external
structure provides a path for noise since it may be
excited by the gearbox at its attachment points and
produce or amplify noise.
7.3 Design
7.3.1 Finite element analysis
The ideal time for minimizing vibration is in the
preliminary design stage. The ability to analyze a
given gearbox and modify its design, based solely
on this analysis, in order to minimize its operating
vibration level requires the use of several finite
element modeling techniques. These analyses
define the excitation due to the gears, the response
of the shaft support system to these excitations, the
manner in which these shaft responses are transferred to the housing through their bearings, and the
response to these various stimuli. In general, the
approach involves the following analyses:

a) Modeling the gearteeth for local dynamicflexibilii and kinematic loading;


d) A detailed finite element model of the static

gearbox structure;
e) An analysis of the modes of the entire system.
7.39 Strain energy considerations
In recent years, a trend to dynamic optimization by
strain energy techniques has evolved. This has
mainly been for the alteration of an undesirable
natural frequency. To understand the technique,
consider that in general each natural mode of the
structure contributes to vibration in proportion to its
amplification factor. Consequently, each mode
whose frequency is in the vicinity of the forcing
frequency would be a major contributor to the
overall dynamic response. In the modal method,
which operates principally on the amplification
factor, the natural frequency immediately above the
exciting frequency is usually increased. One could
also reduce the natural frequency immediately
below the exciting frequency if it is possible
A finite element analysis is first employed in the
modal method to yield a dynamic solution. The
mode shapes are obtained, then the modal strain
energy distribution throughout the structure is found
for any given mode shape whose natural frequency
is to be modified. The strain energies for all
structural elements are obtained and then tabulated
from the highest to lowest. The structural elements
with the highest strain would be the best candidates
for modification of the natural frequency. For
example, in the case of increasing the lowest mode,
the elements with the highest strain density, when
deformed in this mode, would be the best
candidates for modification to obtain a maximum
frequency shii for a minimum material addition
(weight) penalty.
7.3.3 Design parameters
various transmission design parameters affect vibration and noise level, some of the more significant
ones will be addressed in the next subclauses.

When utilizingthis information,the designer should

beaware that they represent trends only. Due to the

AGMA 91%A94

complex interactions which exist in a gearbox, the

results of similar investigations utilizing significantly
different test boxes may vary substantially.

Tooth combinations

The selection of tooth combinations which avoid

potential resonances is sometimes possible. By
varying the selected tooth numbers, the meshing
frequency can be raised or lowered. Obviously, this
must be done within the limitations imposed by
stress and scuffing (scoring) requirements.

Contact ratio

The properties inherent in certain types of gears can

have a large effect on gearbox vibration levels. Accuracy aside, contact ratio (simply stated, contact
ratio is the average number of teeth in contact during each mesh cycle) is one of the most important
parameters which affects gear tooth excitation and
thus noise and vibration level. Virtually all other
gear tooth parameters which affect vibration level
do so largely by virtue of their effect on contact ratio.
For example, decreasing a spur gears pressure
angle generally decreases noise level, however the
same effect can be achieved by extending the
addenda of the higher pressure angle gears slightly
to achieve the same contact ratio as the lower
pressure angle gear. Unfortunately, due to tooth
geometry limitations, this approach can be used
only for relatively small changes in pressure angle.
As noted earlier, the primary exciting force within the
gear mesh is due to the non-uniform transfer of load
between successive pairs of meshing teeth.
Increasing the number of tooth pairs in contact
reduces the total load carried by any single pair of
teeth, thus the dynamic forces generated at engagement and disengagement are reduced. In
addition, as the number of tooth pairs in contact
increases, individual tooth errors tend to average
out, again reducing the dynamic loads generated.
One way to improve contact ratio and thus reduce
noise is to use helical rather than spur gears or spiral bevel rather than straight or Zero1bevel gears.
Generally, when considering vibration and noise
levels, the designer should strive to achieve the
highest contact ratio possible within the constraints
of the system being designed. Since vibration or

noise is seldom the single driving force in the design

of a transmission system, other considerations will
dictate the basic type of gearing to be used. Given
that the box is to be of the parallel shaft type, the
designer may choose from simple spur gears
(contact ratio 1.25 - 1.70), single or double helical
gears (total contact ratio* 1.8 - 5.0, or greater in
some cases), high profile contact ratio (HCR) spur
gears (contact ratio greater than 2), or any of a
variety of special purpose geartypes. The choice of
which specific type of gearing to be used depends
on many considerations, only one of which is its
vibratory characteristic. Tooth shape modification
In many cases, it is desirable to modify the gear
tooth profile shapes so that they are no longer true
involute curves, especially for high load, high speed
drives. This is accomplished by relieving the tip
andor the flank of the profile. This is done to avoid
tooth interference on engagement and disengagement during rotation. This interference is due to
deflections, both within the gear teeth themselves
and in the housing and shafts. If these interferences
are not relieved, the load capacity of the gears can
be seriously impaired. In addition, since these
conditions generate high dynamic loads, vibration
excitation and noise is also increased.
In addition to modifying the tooth profile shape, it is
also often desirable to modii the tooth shape in the
axial direction. This can take the form of crowning or
lead correction. Lead or helii corrections compensate for deflections of the teeth and shafts and also
for the housingsand supports. If pinions have a high
length to diameter ratio (above l/i), and are subjected to heavy loads, they will often have enough
windup to cause heavy end loading if compensation is not provided. This heavy concentration of
load at the ends will contribute to vibration and
noise. Crowning may also be used to relieve end
It must be emphasized that any tooth shape modification is optimum for only a single load level. Vibration levels will increase both above and below this
load level. If the modification is not very severe, the
increase at lower loads may not even be perceptible
but heavily modified gears will be noisy at offpeak

.* Total contact ratio is equal to the transverse contact ratio plus the face contact ratio.


AGMA Qll-A94

loads. For this reason, the actual modification applied will often be a compromise. Gear accuracy
It is usually true that a more accurate gear also has
less vibration, assuming appropriate tooth modifications. If the gear mesh is exciting a resonance
somewhere in the system, nothing short of a
uperfect gear set will substantially reduce the
vibration and noise. In most other cases however,
improving accuracy will reduce the vibratory levels.
This is especially true when relatively low quality
gears are being used but much less so for higher
quality gears. Among the most important characteristics that affect noise and vibrations are involute
hollowness, lead variation, and control of profile
modification. Design of shafts and location of
The design of the shafts which support the gears
and the location of the bearings which support the
shafts are a critical concern for two quite different
reasons. The first is the need to provide the gears
with a support system which maintains their relative
alignment. A misaligned gear system is subject to
higher unit tooth loads which not only degrade load
capacity but also increase the overall noise and
vibration level of the box.
In many cases, especially for high speed, lightweight gears, the support shafts are frequently
sized based on deflection and not solely strength
restrictions. The location of the bearing supports,
as noted earlier, should be chosen so that they
coincide as nearly as possible to node points in the
shaft mode shape. This will ensure minimum
vibration transmission to the housing and thus
minimum noise.

reduce the propagated levels. Often, these ribs can

be spaced so that they also contribute to improved
The second approach to housing design is more
quantitative. Utilizing the results of the finite element analysis, the strain energy density of the entire
housing can be evaluated. With this information in
hand, thoseareas whichare highest in strain energy
density, and thus the most likely candidate areas for
vibration reduction, can be identified and modified. Bearings
Vibration or noise generated by the bearings is generally overshadowed by that due to the gears and
their interaction with the housing.
Vibration created by rolling element bearings can
sometimes be reduced by control of manufacture,
through tight tolerance and surface finish. This can
lead to an increase in bearing cost. Another
approach to reducing vibration may be to select the
quietest bearings from a production lot. Many
bearing manufacturers utilize a sound checkas one
of their quality control devices. The bearing to be
checked is mounted in a standard fixture and run at
a standard speed. its noise level is checked and
those which exceed a given value are rejected. Tooth stiffness
Several factors affect tooth stiffness, such as pressureangle, andtheratiooftooththicknessto height.
In general, the more flexibility the tooth possesses
the lower the dynamic loading will be, and thus vibratory excitation will be minimized. Changing
these parameters in the directions required to improve flexibility may compromise tooth strength. Backlash

First, the housing design can be evaluated qualitatively. The design should be reviewed to ensure that
there is a rigid load path between each bearing
location and the housing mounting points to reduce
deflections. Large flat or gently curved areas on the
housing surface should be avoided since these

Sufficient backlash should be provided to avoid tight

mesh (drive and coast side contact simultaneously)
under all load and temperature conditions. This
backlash should be obtained by tooth thinning and
not by spreading the center distance since the former method does not affect contact ratio while the
latter does. Under conditions of reversing loads or
during periods of zero load operation, excessive
backlash will result in gearbox noise. The loss in
tooth bending fatigue strength due to tooth thinning
must also be considered; however, for most gear

areas tend to vibrate freely (like a drum) when

systems the actual reduction in strength is very

excited. Ribs can be used to stiffen these areas to

small. Design of housing

There are two major approaches to housing design.


AGMA 911-A94

For very high speed gears or wide face gears with

moderate to high speed, an increase in backlash,
above standard values, may be required to reduce
the velocity of the air-oil mixture which is forced
from the tooth space during meshing. Small increases in backlash can result in measurable improvements when the overall noise orvibration level
is dominated by this phenomena. Spur gears are
more susceptible to this phenomena because of the
large trapped volume which exists during meshing. Root clearance
The effect of root clearance is similar to that of
backlash in that a sufficient amount must be
provided to avoid any tip interference with the
mating gear. Additional root clearance (increased
tooth whole depth) above this value will have the
following effects:
- Tooth bending stress will be increased since
the beam length will be increased;
- The smaller fillet radius in the root will also
contribute to increased bending stress; Gear material

For steel gearing, the choice of a specific material
has little effect on thevibration level. If a non-metallic material can be utilized, however, it can have a
significant effect. This is due to two factors. First,
the increased compliance of these materials decreases the magnitude of the dynamic loads and
second, the materials themselves are usually good
dissipative dampers. These factors combine to
make non-metallic materials a good choice when
the operating environment (temperature, load,
speed, lubricant compatibility, etc.) permits. A
secondary benefti of this approach is that such
gears when run in combination with a steel mate
may be run with minimal or no lube. Surface finish
The surface finish on the flanks of gear teeth has
only a small effect on vibration and noise level. Relative influence of factors
Table 12 is a summary of the design approaches
discussed above showing the relative influence
each characteristic has on vibration and noise.
7.4 Analyzing vibration problems

-There will be a reduction in noise caused by the

exit velocity of the air/oil mixture. This is due to
the reduction in the exit velocity provided by the
increased clearance. In most cases, exit velocities are not high enough to be significant;
-There will be an increase in tooth flexibility and
thus an improvement in load sharing and error

7.4.1 The excitation mechanism

The initial step in analyzing vibration problems is
usually to define the excitation at each gear mesh.
The factors described in table 12 can all contribute
to this excitation.
Table 12 - Potential influence of design
features on noise and vibration
Design feature Gear mesh phasing

Sound transmission is accomplished through the
propagation of waves which can be considered
sinusoidal. The addition of two sine waves with
identical frequencies which are separated in phase
would produce a third wave of the same frequency
but with a third amplitude. While gear noise may not
be exactly sinusoidal it can be phased. Studies of
such phasing have proved that a significant reduction in noise level can thus be achieved in applications having multiple load paths. Examples of
methods to use phasing to reduce noise and
vibration are: tooth selection in multiple path
systems, unequal spacing of planet pinions, and
change of double helical gear intersection point.

Contact ratio
Load intensity
Tooth profile modification
Bearing location
Bearing type
Housing design
Tooth numbers
Tooth stiffness
Root clearance
Surface finish


AGMA 91%A94

A gear system transmitting power is also susceptible to torsional vibration. The inertia may be concentrated as in the body of a gear or distributed as in
the shafting. Similarly, the elasticity may be concentrated as in a coupling, or it may be distributed with
the inertia in the shaft sections. The excitation may
come from externally applied pulsating torques or
from a fluctuating resistance to the steady rotation.
However, in a geared system there is also an excitation due to displacement which comes from the imperfect transfer of motion (as described in 7.1 .l) between the meshing gears. Having defined the gear
tooth excitations, a modified Holzer analysis can be
used to calculate the dynamic tooth forces [9].
7.4.2 Response characteristics Shaft response
Once the dynamic forces at the gear mesh have
been calculated, they can be applied to an analytical
model of the shaft system. The results are dynamic
forces at the shaft support (bearing) locations and
vibration mode shapes of the shafts. The objective
of the shaft analysis is to reduce the shaft response
in general and to change the shaft geometry such
that the bearings are located close to node points for
the critical modes. Gear and housing response
The final step in the analysis is the evaluation of the
response of the gears and housings to the excitations transmitted to them. This task is usually
accomplished through the use of a finite element
program. The responses are obtained by applying
the appropriate forcing functions at the tooth contact
points in the case of the gears or at the bearing
locations within the housing. In the development of
the model, the parts are supported at their normal
mount points by appropriate constraints [lo].
7.4.3 Test methods
Many different methods of testing are used to determine vibratory characteristics. Each method is designed to answer certain kinds of questions, but no
one test will resolve all vibration or noise issues.

loading can be applied. This method offers the

advantage that the parts are subjected to the
operating dynamic conditions, so that results will
closely resemble field experience. Siren
This is usually a non-running test in which the part
to be tested is excited by variable frequency air
pulses. A stream of pressurized air is passed
through a rotating serrated disk whose speed can
be varied over the frequency range of interest. The
air is directed at a location which would excite the
mode shapes of interest. Since several diierent
disks can be used with differing hole patterns,
frequencies up to about 20 kHz can be investigated.
The response can be determined either by using a
camera if there is sufficient motion or with strategically located strain gages. Since this method
applies a low dynamic force to only one specific
component, it is not usually practical to simulate the
response of a complete gearbox by this technique. Impact
This is also a non-rotating test in which the
component to be tested is supported in a manner
which will not affect its frequency response, and
then is struck with an impact device, typically a
hammer with a dynamic force gage, selected for the
frequency range of interest. The response is
measured with accelerometers or strain gages
placed at critical locations. By impacting the part in
various locations and monitoring the responses, it is
possible to determine natural frequencies and
mode shapes. However, since it is a static test, any
change in frequencies or damping characteristics
due to the effect of rotation is not included. Shaker
This type of test is similar to the impact test except
that an electrodynamic shaker head is mounted so
as to impart a force to the component being tested.
The force can be varied in frequency so that the
response can be determined at critical frequencies,
either with accelerometers or other transducers.
Since it is a non-rotating test, the same cautions
apply as with impact testing. Excitation methods Response measurement .I Running tests Strain gages

If the parts to be tested can be operated under

realistic conditions, simulated or actual service

When one of the parameters to be studied is the

stress in the part during vibratory response, strain


AGMA 911-A94

gages can be applied to locations which are

expected to show high stresses. The gages are
used to determine both the frequency response and
the operating stresses. For the results to be useful,
the excitation method must simulate operating


Since the application of strain gages can be diiicult

and time consuming, accelerometers are often
used as the transducers. Similar results with
respect to frequency response and mode shape can
be obtained except with no information relative to
stress. The availabilii of small, lightweight accelerometers has made this technique very popular. Powder


This technique for determining modes and frequencies involves coating the part with a special powder
and exciting it with a variable frequency shaker. By
observing the patterns developed in the powder at
the parts natural frequencies, the modes for those
frequencies can be determined. It is also a nonrotating test, so effects of rotation are not included. The
part is normally excited during the observation by a
shaker, since the air siren would interfere with the
patterns. Holography
This is a laser based optical technique of observing
and photographing mode shapes. The components
are excited, usually by a shaker, at the selected
The holographic camera records a
hologram of the part which is vibrating during the
exposure. The resulting image is covered with light
and dark fringes which are contour lines related to
the vibration amplitude. From this fringe pattern, the
mode shapes can be determined and, if desired, the
pattern can be quantitatively evaluated to determine vibration ampliiude at any point.
The approach is most useful for non-rotational
determination of modes, since obtaining a fringe
pattern on rotating parts requires an elaborate
setup. Table 13 lists some common combinations of
these techniques and shows the characteristics of
7.5 Vibration/Noise



In cases where hardware is in existence before a

noise or vibration problem has been recognized and

must center
around modification to the existing hardware. Several of the more effective methods will be discussed.
Table 13 - Vibration


Test objectives

Strain gage
Strain gage
Strain gage













Isolation means that excitation forces are prevented

from being transmitted from one part of the system
to another. This interruption yields a reduction in the
transmitted forces. It is most common to isolate
either the transmission from its supporting structure
or the gear rim from its blank and hub. The former
approach is generally easier to accomplish and is
therefore the most common.
7.5.1 .l




Vibration isolators are often used to control the

transmission of high frequency energy across the
mounting points to the supporting structure. This
method involves the use of elastomeric mounts between the transmission mounting pointsand its supporting structure. Reductions as high as 10 dB can
be obtained through the use of rubber mounts having a static spring rate equal to that of the supporting
structure at the gear box attachment point.

Gear rim isolation

A second method of reducing vibration and noise is

to s olate the gear rim from its hub with an elastomeric insert. The elastomer will reduce blank excitation and, consequently, shaft and housing exci-


AGMA 911-A94

tation are also reduced. One of the difficulties with

this approach is the limitation on torque transmission imposed by the strength of the elastomer. The
effect of the lubricant on the elastomer and the accuracy of the position of the gear with respect to its
mate may also limit its use to non-critical, low load
situations. In those systems where load capacity is
secondary and vibration or noise is primary, this
method has been shown to work quite well.

Dissipation Externally




Application of damping material to the surface of a

structure by spraying, gluing, plasmaflamecoating,
etc., may be effective. Since most of the straining
action (and thus energy conversion into heat) will be
confined to the layers closest to the structure, thin
layers of damping material will be more cost and
weight effective than thick layers.
Objections to the use of damping materials applied
to the exterior surface of transmission housings
. include greater cost, added weight, concealment of
cracks, and heat retention. Each of these must be
traded off against the potential benefits.
There are many methods by which dissipative
damping may be achieved. The simplest of these is
to fill a hollow gear shaft with an elastomeric
material, or to provide constrained layer damping.
Dissipative damping is effective when constrained
layer damping is applied to gears or housings with
large flat surface areas which tend to ring when
excited. Damping
_ - rings
Rotating parts such as bevel or spur gears have
many natural modes of vibration which can be
excited within the operating range.
It is often
impractical to redesign the part to avoid all of these
frequencies, so some means of reducing the
response is the next best solution. The use of
damping rings is quite common in this situation.
These can take the form of special snap rings
installed in grooves in the gear rim, or more complex
rings designed specifically for the purpose. Figure
13 shows an example of a typical damping ring for a
spiral bevel gear rim. The location of the grooves
and the shape of the ring are determined by the
mode shapes of concern and by the need for the


frictional damping to occur at the interface of the

rings and grooves. Care must be taken to avoid
having the rings become a wearing part and thereby
introducing an additional mode of failure.

Figure 13 - Typical damping



Screens or barriers

A well constructed screen can shield a high frequency source and result in a significant noise level
reduction. However, if the gear drive is operating at
avery low speed, it will produce IowfrequWcy noise
and the gap around the screen can act as a new
source and radiate sound into the space with little
attenuation. Some noise will still enter the working
area via reflection from the adjacent walls as well as
by transmission through the-screen itself. In this
case, the use of absorbent materials, in conjunction
with a screen, can prove quite effective.



Sound absorbing materials serve the purpose of

minimizing reflection of sound waves. They are not
very effective at blocking the path of sound transmission. For this reason, they are usually used to
alterthe sound characteristics of an area, and in this
manner, reduce the sound pressure level within that
area. This type of material is not normally used
alone as a barrier between a source and a receiver,
since the sound pressure level would be reduced
very little at the receiver. Some factors affecting the
absorption characteristics of a material are mass,
surface condition, pore size and structure, flow
resistance, thickness and the frequency of impinging sound. Typical materials used are glass fiber
and certain types of polyurethane foam.
7.5.5 Enclosures
Reduction of airborne noise by the use of enclosurescan be effective, but the noiseattenuation that
can be achieved is dependent upon the complete-

AGMA 911-A94

from the stress index prediction. The reasons for

disagreement between actual stresses and stress
index numbers can be caused by any number of
factors including the following:

ness of the enclosure. The noise reduction limitation in the speech frequency range with typical
acoustical enclosures and seals is about 25 dB, with
up to 35 dB obtainable through use of improved seal
configurations. Further reductions in noise level, up
to 50 to 60 dB, can be achieved with fume-tight
enclosures, such as those employed in some commercial helicopters and in some commercial transport aircraft engine installations operating today.

- Size effects;
- Fine pitch vs. coarse;
- Blank configuration;
- Rim;
- support;

Practical enclosures are limited in noise attenuation

by unavoidable sound leaks in seams and access
doors. Not only do these enclosures impose
considerable weight and maintainability penalties,
but they do not reduce the harmful effect of the
which contribute to
material fatigue and fretting at joints.

A summary chart showing the relative effectiveness

of the various methods discussed

given in Table 14.

in this clause is

Table 14 - Noise and vibration

reduction techniques
I Relative effectiveness

Mounting isolation
Rim isolation
External damping
Internal damping rings
Absorbent material





8.1 Introduction

for fatigue

The conventional AGMA method of bending stress

calculation is in reality a stress index method of
design. Production designs based on this method
have countless hours of successful field operation.
The allowables are based on this successful field
experience and successful/unsuccessful
test experience. When the stresses are determined


by strain gaging gear teeth, it is

found that the actual stresses may be quite different

Thus the designer faces the dilemma of whether or

not to use the stress index method or to conduct a
more complex analysis, such as finite element
analysis. This is not to say that the classical method
is inaccurate for it can be shown that the final
application factor by the stress index method and
FEA are in agreement if the proper values for the
factors have been chosen.

8 Load capacity

Dynamic effectslresonance;
Temperature effects;
Loading conditions/prelubing;
Deflections of teeth, shaft, housings.

The conventional AGMA method permits calculation of the maximum tooth tensile stress. The
minimum stress is assumed to be zero. Thus the
vibratory and steady stress is taken to be l/2 of the
maximum tensile stress. The true minimum bending stress in the gear is usually negative and occurs
prior to the initial point of contact for a driving gear
and after the final point of contact for a driven gear
(see figure 14). Afinite elementanalysiscan
the maximum and minimum stress by conducting a
series of static solutions of stress as the pinion and
gear are incrementally rolled through mesh. The
stress allowables for the finite element method are
generally taken to be the material allowables found
from material testing or from actual gear teeth test
A design approach that has been employed is to
conduct the design using the classical AGMA stress
index method and then to conduct an optional finite
element analysis to refine the design. A digital computer analysis using AGMA equations can quickly
and efficiently compare many preliminary designs

whereas the finite element method is very tedious

and time consuming.

On the other hand, the FEA

39 39

AGMA Qll-A94


Maximum stress, thick rim (AGMA stress index)



Minimum and maximum stresses, thin rim (FEA method)

Figure 14 - Different methods for determining tooth root stress
can more readily account for variables such as
suppottstiSfnessand temperature effects. Accounting for these factors permits the load distribution to
be determined.
In aerospace gearing where the gear rims are
designed to minimum thickness, the gear rim and
web can have a profound effect on the bending
stress. Thin rims make the calculated stress
number too low, giving an optimistic estimate of life.
As the thickness of the gear rim is reduced, there is
a point where the stress in the gear rim itself is
higher than the tooth root bending stress. This is
generally not a good design condition because
cracks may propagate through the gear rim instead
of through the tooth. It is of course not desirable to
have any cracks which propagate, but assuming a
crack were to be present, it is desirable to control the
direction of propagation.
Cracks through the gear rim should be avoided because they can cause sudden and catastrophic loss
of the mesh without warning since chips are usually
not generated by this type of fracture. Hence the
stress in the root of the tooth should be lower than
the tooth bending stress which is usually maximum
at a position near the TIF. Stresses in the gear rim


and web can be calculated by FEA. The user should

satisfy himself that the gear blank construction is
representative of the basic theory embodied. Figures 14 and 15 illustrate the above discussion.

through tooth

Maximum tooth
rim stress

Direction of propagation through rim
Figure 15 - Directions of crack propagation
in gear teeth

8.12 Design criteria

Prior to beginning the gear analysis several design
parameters must be established. In aerospace
gearing, the design is usually based on three
compression (Hertz stress), and scuffing (scoring)
resistance. Each of these conditions must be


checked independently for adequacy of design. For

aerospace designs which incorporate condition
monitoring, bending tooth breakage probably has
the most severe consequence of the three failure
modes whereas pitting and scuffing are durability
type failures.
Durability type failures can cause gearbox removals
but are not catastrophic unless the gearbox is permitted to operate in that condition for a sufficient
time period in which the surface breakdown degenerates to the point where the mesh is lost. Pitting
and scuffing are a phenomenon of the mesh while
bending is individual to the pinion and gear, therefore both members must be analyzed for bending.
Optimization of scuffing and a bending stress balance or life balance between pinion and gear can be
achieved by varying addendums and dedendums
and/or tooth thicknesses of both members.
In aerospace designs, it is desirable to achieve a
balanced tooth design. That is to say that the life in
bending, contact, and scuffing resistance should be
optimized. This is not always possible for a number
of reasons, but the gear designer generally has
control of basic parameters such as the number of
teeth, diametral pitch, and center distance. In
general, the gearset is designed for Hertz stress (for
single external mesh) as follows:
315ooP (V&+ 1)3


A- nP G
6.923 x lo8 P (mG+ 1)3

62F =


nP mG

From equation 3.1 of ANSVAGMA 1012-F90,


. ..(lO)

and if F is set equal to d, and substituted in 9, the

cubic equation can be solved ford, (and F), where

is center distance, in (mm);


is operating pitch diameter of the pinion,

in (mm);

is face width, in (mm);

is contact load factor for pitting
resistance, lb/in* (N/mn?);

is transmitted power, hp (kVV);

is gear ratio;

np is pinion speed, rpm.

NOTE:K= 500400 istypicalforcontinuousoperation
of carburizedaerospacegearing.
The teeth and diametral pitch can be selected for
bending. The gearset scuffing resistance is then
8.1.3 Design life considerations
Of primary importance in the preliminary design are
reliability requirements. The life can be calculated
using Miners cumulative damage theory (see
Appendix B of ANSVAGMA 2001-B88 for explanation of Miners Rule). The gear designer must have
a spectrum of estimated operating power, rpm, and
percent time. This usage spectrum should be
established by analysis of the operating environment and intended use of the gearbox being
designed. Even in the case of relatively slow speed
drives, the cycle build up is very rapid. Unless
operation at the highest powers is for very limited
durations, finite life design becomes impractical.
On the other hand, if the high power condition is truly
transient in nature and the rpm is relatively low, finite
life design can achieve reduced size and weight. To
conduct a finite life design, material stress vs. cycle
curves must also be known (see ANWAGMA
2001-B88 for allowable stress and life factors). If it
is desired to have unlimited life, the gearing can be
designed for limit load or the maximum load
expected in service.
8.2 Spur, helical and bevel gear tooth breakage
and surface durability
The bending strength of gear teeth is a fatigue
phenomenon related to cracking at the tooth root
fillet. See ANSVAGMA 110.04-1980 for a more
complete discussion.
The basic theory assumes that the tooth is a
cantilever beam fixed at its base (the gear rim) with
a reduction in tensile stress caused by the
compression of the radial component of the
transmitted load. A stress concentration factor is
applied because the tooth is subjected to alternate
cycles of maximum to zero stress. Load sharing
between adjacent teeth in contact is calculated.
The pitting of gear teeth is a fatigue phenomenon
caused by repeated application of surface com-


AGMA 91%A94

pressive stresses as the teeth roll through mesh.

Initial pitting and destructive pitting are illustrated
and discussed in ANSVAGMA 110.04.

SC = s

w* G


cnl Cf

. ..(12)

8.2.1 Fundamental bending stress formula
(Refer to ANSVAGMA 2001-888 for spur or helical
gears or ANSVAGMA 2008-A86 for bevel gears)
5 --v wt









s _ Wt Kca 1 5


. ..(ll M)






is calculated tensile bending stress number

at the root of the tooth, lb/in* (NImrr?);
is tangential tooth load, lb (N);
is external application factor for bending
strength. This factor takes into account the
effect of any externally applied load in excess of the nominally applied torque load;
is the rim thickness factor for spur and helical gears (use 1 .Ofor bevel gears);
is dynamic factor for bending strength. This
factor takes into account the effect of gear
tooth quality, as related to speed and load;
is size factor for bending strength. This
factor takes into account nonuniformity of
material properties;
is load distribution factor for bending
strength. This factor modifies the rating
equation to reflect the non-uniform distribution of load across the face of the tooth;
is tooth lengthwise curvature factor for
bevel gear bending strength (use 1 .O for
spur or helical gears);
is tooth geometry factor for bending
is diametral pitch, in-l;
is module, (mm);
is face width of gear for which bending
stress is desired, in (mm).

8.2.2 Fundamental contact stress formula

Refer to ANSVAGMA 2001-888 for spur or helical
gears or ANSVAGMA 2008-A86 for bevel gears.



is contact stress number, lb/in* (N/mm2>;

is elastic coefficient, W
is application factor for pitting resistance;
is dynamic factor for pitting resistance;
is size factor for pitting resistance;
is operating pitch diameter of pinion, in
is effective or net face width of narrowest
member, in (mm);
is load distribution factor for pitting
is surface condition factor for pitting

is geometry factor for pitting resistance.

8.2.3 Dynamic factor
A dynamic factor is applied to the tooth bending and
contact stress to account for increased dynamic
tooth load. lnertias and spring rates of the
transmission system influence the dynamic tooth
load as does operating speed compared to resonance speed. Gear tooth tolerances have a large
effect on dynamic load. As the gear tooth rolls
through mesh, teeth are engaging and disengaging
as a function of the tooth geometry. For example in
a conventional mesh, there are alternate cycles
where the load is carried by varying numbers of
teeth. In the worst case, since the spring rate of the
mesh is considerably different when one tooth is
meshing and when two teeth are in mesh, the
tangential deflections are different at different times
during the mesh cycle.
This change of mesh stiffness with time causes the
mesh point tangential deflection to vary with time at
the tooth mesh frequency. The resultant change in
tangential deflection caused by a change in velocity
from the theoretical constant velocity is sometimes
referred to as transmission error or TE. Transmission error induces a vibratory forcing function at the
mesh point and is the root cause of dynamic tooth
load and noise.
Profile modifications on gear teeth are required to
accommodate tooth mesh deflections and other

AGMA 911-A94

variables. Modifications can also have an effect on

tooth spring rate. The modification specified is
optimum for only one load condition, usually some
high load, high speed operating condition. When
operated at other load conditions, the gearset may
produce higher dynamic loads. Excessive profile
modification can actually reduce mesh contact ratio
and introduce error in action.
A complete discussion and calculation of transmission error and resultant dynamic load is beyond the
scope of this guide. A simplified dynamic factor can
be obtained by referring to ANSVAGMA 2001-S88
for spur or helical gears or ANSVAGMA 2003-R%
for bevel gears.
8.2.4 Siie factor
Size effect factor has been established to account
for detrimental effects as the volume of material in
the gear set increases. This can be thought of as a
chain analogy. The more links in the chain (units of
stressed volume), the higher the probability of a
defect in one of the links. For aerospace spur and
helical gears the gear size has not been established
to have a detrimental effect and a factor of 1.O is
used. For bevel gears a size factor is used and can
be determined by referring to ANSVAGMA
8.2.5 Load distribution factor
The load distribution factor accounts for
non-uniformity of load across the lines of contact on
the teeth. The deviation from the theoretical uniform
load is caused by operating deflections and
dimensional allowances.
Deflections along the teeth are caused by the following:
- elastic deflections of the gear elements;
- shaft deflections;
- bearing deflections;
- housing deflections;
- foundations which support the gear
-thermal expansion from non-uniformity of
- differential thermal expansion from
different materials;
- centrifugal effects.

Dimensional allowances that affect load distribution

are as follows:
- accumulation of tolerances;
- alignment of the axes of rotation;
- bearing clearances;
- profile accuracy;
- lead;
- crowning.
For a particular application, each of the above
influences should be evaluated as to its effect on
load distribution. Modem finite element methods
can be used if the pinion and gear are modeled as
separate parts and the tooth load induced by torque
applied to the shafts through the use of gap
elements or with three dimensional contact modeling. The finite element technique is suitable for
calculating load distribution because all of the
factors which influence deflection and manufacturing deviations can be evaluated.
89.8 Geometry factor, J
The geometry factor is used to calculate gear tooth
bending stress. The geometry factor takes into
account bending of the tooth as a cantilever beam
as well as the compression caused by the radial
component of the normal tooth load. The maximum
stress occurs where the load is at the highest point
of single tooth contact for a conventional low contact
ratio gear tooth. For high contact ratio gears a load
sharing factor is applied to account for distribution of
the transmitted load among the teeth.
Internal gears are often neglected in texts but are
important in aerospace design because of heavy
reliance on planetaries which have an internal ring
gear. Annex A shows a procedure for calculation of
the geometry factor for an internal gear mesh. The
procedure is suitable for calculation of either
internal or external gears or mates. The equations
are generalized by the use of signed integers
which are +l for external gears and -1 for internal
Helical gear geometry factors are based on a mean
normal tooth section and also account for load
sharing. After the mean normal section is established, the geometry factor iscalculated in the same
manner as for a spur gear. See Annex A for spur
gear geometry factors.
Further data on the
calculation of spur and helical gear geometry


AGMA 911-A94

factors is shown in AGMA 908-B89. For bevel gear

geometry factors, see ANSVAGMA 2003-A86.
8.2.7 Allowable bending stress
The allowable bending stress is calculated as an
allowable stress index number and represents the
allowable stress when the stresses are calculated
by the procedure outlined above. The allowable
stress number is a function of the material and heat
treatment used as well as the desired life,
temperature of operation, and reliability desired.
The relationship is given by:





sWt is working bending stress number, lb/in*


is allowable bending stress number, lb/in*



is life factor for bending strength;


is temperature factor for bending strength;


is reliability factor for bending strength.

The reliability factor accounts for the normal statistical variations found when materials fracture. From
the analysis of probability, for a normal distribution,
the reliability constant is given as:
-Rel = 1-nv


...(i 4)

is reliability constant;
is number of standard deviations;
is coefficient of variation = standard

It has been found experimentally that for steel, the

coefficient of variation is approximately 10% (v =
0.1). This number can be used for spur and helical
gears but the variation in spiral bevel gears has
been found to be higher because of problems
encountered in shimming or other dimensional
variables which influence gear tooth patterns. A
standard deviation of 0.156 has been calculated for
bevel gears using data from a large number of test

16. For highly reliable aerospace design, a

reliability of 3 standard deviations has been used in
the past (or 3o). This results in a reliability of
0.99875 and a reliability constant of 0.7. Thus the
working 3o allowable is found by multiplying the
mean stress by 0.7.
From this discussion it is seen that there can be
different allowable design stresses for the same
design and that each stress will have a
corresponding reliability associated with it. It is up to
the designer to establish reliability goals before the
design begins so that allowable stress can be
determined. The allowable stress numbers shown
in ANSVAGMA 2001-B88 are based on a reliability
of 1 failure in 100 at 1Oscycles. This corresponds to
n = 2.326 (number of standard deviations
corresponding to 99% reliability). Thus for a spur or
helical gear with v = 0.1, the reliability constant for
99% reliability is 0.7674 and for a bevel gear with
v = 0.156, the reliability constant is 0.6371. To
determine the reliability factor for multiplication of
the stress index:




. ..(15)



is reliability constant for 99% (0.7674

for spur or helical gears and 0.6371 for
a bevel gear);

Rdesired = 1-nv = reliability constant for desired

The allowable bending stress numbers and factors
are shown and discussed in ANSVAGMA
2001-888 for spur and helical gears and in ANSI/
AGMA 2003-A86 for bevel gears.
Reliability (or probability) is related to the number of
standard deviations by the term:
Rel =
Unfortunately, there is no closed form solution to
this integral. However, it may be solved by a

and field bevel gearfractures. A plot of reliabilityvs.

numerical approach. For a numerical approach to

number of standard deviations is depicted in figure

calculate reliability, knowing the numberof standard


AGMA 91 l-A94

deviations, n:

Rer= 1 (...17)

8.3 Spur, helical, and bevel gear scuffing

(scoring) - flash temperature index


bending stresses. The allowable contact stress

numbers and factors are shown and discussed in
ANSVAGMA 2001-B88 for spur and helical gears
and in ANSVAGMA 200%A86 for bevel gears.

= l+

8.3.1 Fundamental flash temperature



Cl = 0.319381530

The fundamental scuffing (scoring) formula is

derived by the method proposed by Blok which
calculates a temperature at the gear mesh point as
the gear bulktemperature added to the temperature
rise in the mesh. The temperature rise in the mesh
is referred to as the flash temperature index.
Appendix A of ANSUAGMA 200%B88 gives a
detailed discussion of the flash temperature index
method for spur or helical gears, but there is no
generally accepted AGMA method for bevel gears.

c:! = -0.356563782


c3 =

c4 =- 1.821255978


n = number of standard deviations

(horizontal axis of figure 16)


reliability (vertical axis of figure 16)

By the above method, Relwill be accurate to 4 decimals.

8.2.8 Allowable contact stress
The allowable contact stress is a function of
parameters similar to those used for allowable
bending stress number and in addition includes a
hardness ratio factor. The relationship is given by:

sac cL cH




For bevel gears, the flash temperature index

method outlined by Gleason has been widely used.
Both the Gleason and AGMA methods are based on
Blokand Kelley and the general form of the equation
is as follows:
tc = tM + tfl
tc = t&f + t,

. ..(19)
c 17.78

. ..(19M)

tc is contact temperature in F (C);
?M is bulk temperature in OF(C);

is flash temperature in F (C).



is working contact stress number in pounds

per square inch (megapascals);
is allowable contact stress number in
pounds per square inch (megapascals);
is life factor for pitting resistance;
is hardness ratio factor for pitting resistance;


is temperature factor for pitting resistance:

is reliability factor for pitting resistance.

To calculate CR , use the equations forKR shown in

8.2.7. For bevel gears, a lower coefficient of variation has been observed for contact stresses than for

8.33 Film thickness method

The film thickness method is based on the work of
Dowson and Higginson. The minimum oil film thickness is calculated under load and is divided by the
composite surface roughness which is an average
surface roughness of the pinion and gear. When the
ratio of film thickness to surface roughness, h, is 1.O
or greater, the surface asperities cannot touch and
there is no scuffing (scoring) or wear. When the
pitch line velocity of the mesh is low [PLV < 1500 ft/
min (7.6 m/s)], acceptable values of surface film
thickness and h are achieved through the use of
high viscosity oils. The calculations are summarized in ANSVAGMA 2001-B88 in Appendix A.


AGMA 91%A94







Figure 16 - Reliability versus number of standard deviations


AGMA 911-A94

9 Gear materials and heat treatment

Aerospace gears are manufactured from appropriate materials to perform under the imposed operating conditions for the life of the gear application.
Gear material specifications are determined by the
requirements of the application for mechanical
properties, material quality, dimensional stability,
hardenabilii, and manufacturing characteristics.
Heat treatment is specified to achieve the required
mechanical properties of the gear materials.

the material must endure. Mechanical properties of

the material determine the allowable stress levels for
the application.

9.2.1 Hardness
The strength properties are closely related to
material hardness for ferrous materials. Hardness
indicates the strength of the material and resistance
to tooth bending failure. Surface hardness provides
resistance to gear wear, pitting and scuffing.

Ferrous and non-ferrous materials are used for

aerospace gearing. Steel alloy materials are used in
most primary gear applications for strength and
durability. Stainless steels are used in special
applications for corrosion resistance. Bronze alloys
are occasionally used in worm gearing for wear
resistance and reduced friction coefficient.

9.2.2 Fatigue strength

9.1 Class and grade definitions

9.2.3 Tensile strength

Aerospace gearing is divided into three classes depending on the nature of the intended application.

Tensile strength is the maximum possible tensile

stress before fracture occurs. Tensile strength is
considered as a mechanical property used for relative comparison of materials. Hardness is recommended in lieu of tensile strength for specifications
in gear manufacture.

Gear materials are specified by grade of metallurgical control factors and allowable stress numbers according to ANSVAGMA 2006B88 and ANSVAGMA
2003-A88 The three grade numbers are: Grade lmoderate quality, Grade 2 - superior quality, and
Grade 3 - Premium quality.
Gear application classes are described with the material grade numbers that are used as follows:
a) Typically Grade 3 material is used in main
drive system components, where failure of a
gear could result in the loss of the vehicle or
endanger operating personnel or passengers;
b) Typically Grade 2 or 3 material is used where
failure of a gear may render the primary system
inoperative, but where a secondary system can
be engaged to perform the same function;
c) Typically Grade 1 or 2 material is used where
failure of a gear may affect mission capability or
an auxiliary system but will not result in the loss
of a vehicle nor endanger operating personnel
or passengers.

Contact and bending fatigue strengths predict the

number of cycles that a gear can endure at a given
stress level before surface pitting or tooth fracture
occurs. Contact and bending fatigue strengths are
influenced by hardness, microstructure, material
cleanliness, surface conditions and residual stress.

9.2.4 Yield strength

Yield strength is the maximum possible stress
before permanent deformation occurs.

9.2.5 Toughness
Toughness is a measure of ability to absorb impact
energy and is important for high impact or low temperature applications. Brittle fracture may occur in
high strength materials as a sudden failure or rapid
crack propagation below the tensile strength due to
low fracture toughness properties. Toughness of
steel gearing is adversely affected by the following
- Low temperature;
- Improper heat treatment causing defective
- High sulphur content;

9.2 Mechanical properties

- High phosphorus and embriiling grain bound-

The particular design configuration and duty cycle

ary precipitates;

loads of the application determine stress levels that

- Non-metallic inclusions ;


AGMA 91%A94

- Large grain size;

- Absence of nickel alloying element ;
- Stress concentrating notches, fillets, tool
marks or rough surface finish.
9.3 Cleanliness
Cleanliness determines the extent of homogeneous
material properties. Alloy steel manufactured with
electric furnace practice is commonly vacuum
degassed, inert atmosphere (argon) shielded and
bottom poured to improve cleanliness and reduce
gas content. Reduced non-metallic inclusion content improves transverse ductility and impact
strength. Further refinement by vacuum arc remelting o/AR) or electroslag remelting (ESR) reduces
gas and inclusion content more for improved fatigue
strength. Cleanliness requirements of alloy steels
are controlled according to AMS 2300 (premium
aircraft quality steel) and AMS 2301 (aircraft quality
9.4 Heat treatment
Most wrought ferrous materials used in aerospace
gearing are heat treated to meet. hardness and
mechanical property requirements. Gear blanks are
generally annealed and normalized to produce
homogeneous microstructure for uniform machinabitii and improved response to subsequent heat
treatment. Quenching and tempering increase
material hardness and strength properties. Case
hardening increases surface hardness and strength
while maintaining a softer core for toughness.

The surface is often required to have high hardness

and high strength, while the core is required to be
tough for impact resistance. A correctly hardened
steel gear will consist primarily of tempered
martensite. Furthermore, a gas carburized gear will
contain a high carbon tempered martensite case
with a low carbon tempered martensite core.
Excessive case carbon content can result in
retained austenite if not transformed and can also
lead to the formation of undesirable carbide
networks at grain boundaries. Improper hardening,
such as quenching, can result in undesirable bainite,
free ferrite or pearlite, in the from of banding.
9.6 Hardenability
Hardenability of steel is the property that determines
the hardness gradient produced by quenching from
the austenitizing temperature. The asquenched
surface hardness is dependent primarily on the
carbon content of the steel and cooling rate. The
depth to which a particular hardness is achieved with
a given quenching condition is a function of the
hardenabilii due to carbon and alloy content of the
9.7 Dimensional stability
Improper heat treatment processes cause dimensional distortion and possible cracking. Distortion is
caused by mechanical and thermal stresses with
phase transformation during quench and tempering
due to variations in: 1) section thickness, and 2) the
duration at transformation temperature.
9.6 Pre-machining

9.5 Microstructure
The microstructure is the material structure
observed at 100X or higher magnification and
reveals the constituents of the material. The
constituents include, but are not limited to,
martensite, ferrite, pearlite, and bainite. The
microstructure also reveals grain size, carbides,
carbide networks, and retained austenite. These
constituents are a result of the heat treat process
and can help determine if a heat treat process was
done correctly or incorrectly. lt should be noted that
on a cross section, the microstructure near the

surface may be dierent from the microstructureof

the core.

stock removal

A specified minimum thickness. of surface stock

should be removed from ferrous gear forgings to
eliminate decarburization, seams, and other surface
9.9 Ferrous gearing
Ferrous materials for aerospace applications are
primarily wrought alloy steels. Wrought steels are
mechanically worked to form round stock, flat stock,
or forgings. Anisotropic mechanical properties

(tensile ductilii, fatigue strength, and impact

strength) vary according to the direction of hot
working or inclusion and grain flow. Improved steel
cleanliness improves the transverse and tangential

AGMA 911-A94

properties of forged steel that approach, but do not

equal, longitudinal properties.

9.11 Material grades and heat treatment

Common gear materials and heat treatment used in
aerospace applications are shown in table 15.

Other ferrous materials may include H-series tool

steels or austenitic, marten&c, and precipitation
hardening stainless steels for special requirements.
Powdered metal materials may be used in certain
non-critical applications.

9.12 Gear surface hardening

9.10 Non-ferrous gearing

Copper base (bronze) gears are used in worm gear
applications with steel worms to improve wear
resistance and reduce the coefficient of sliding
friction. Manganese bronze and aluminum bronze
have higher strength than phosphor or tin bronze

Most aerospace power gears are produced with

heat treated alloy steels and surface hardened to
provide tooth bending strength and resistance to pitting and wear. Carburizing is the primary method of
surface hardening gear teeth. Nitriding is an alternate surface hardening process specified where
carburizing and quenching would cause excessive

Table 15 - Typical aerospace gear materials




Typical hardness



Alloy St-L .
AISI 9310










Main drive,
Accessory, actuators
Main drive,
High temperature
Main driie,
High temperature
High temperature
High temperature
Accessory drive
Accessory drive
Accessory drive
Accessory drive



HP 9-4-30



Nitralloy 135M
Niiralloy N
AISI 4340
AISI 4340
Stainless steel:
PH13-8 MO
Custom 455





Accessory drive


Worm gear



C-H = Carburize and harden

TH-N = Through harden and nitride
5: Through harden
= Precipitation harden
= Solutionheat treat



Rockwell hardness scales (HRC and HRB) are shown for

direct comparison only. In general, those scales are not
specifically recommended for measurement where other,
more accurate hardness scales are commonly used.


Proprietary material designation.


AGMA 911-A94

However, nitriding does not increase tooth bending

strength as much as carburizing. Other methods of
selective direct hardening may be used for less critical gear applications with medium carbon alloy
The minimum depth of surface hardened case in
gear teeth is required to resist: (1) sub-surface
shear stress caused by tooth contact pressure and
(2) root fillet tensile stress caused by tooth bending.
Minimum required case depth is typically specified
as the greater value of: (1) two times (2x) the depth
of maximum sub-surface shear stress at the pitch
line, or (2) 10 percent of the tooth thickness to
accommodate sub-surface fillet tensile stress.
Refer to ANSVAGMA 2001-B88 for carburiied or
nitrided case depth recommendations.
The case must not be so great as to result in brittle
teeth tips and edges, or high residual tensile
stresses in the core. Maximum case depth at the
tooth tip should be limited to 56 percent of the tooth
top land thickness when possible.
The effective case depth for carburized and
hardened gears is defined as the depth below the
surface at which the hardness is 50 HRC equivalent
by conversion from microhardness.
The effective case depth for induction hardened
gears is defined as the depth below the surface at
which the hardness is 10 HRC points below the
specified minimum surface hardness.
For nitrided gears, case hardness is specified as total case depth, and is defined as the depth below the
surface at which the hardness has dropped to 110
percent of the core hardness. The practical limit for
maximum nitrided case depth is 0.040 inch (1 mm).

After carburizing the gear is cooled slowly in the

carburizing medium or protective atmosphere to
control decarburiiation. Next the gear may be
reheated to 1200F (649OC)for subcriiical anneal to
reduce sudden changes in carbon content across
grain boundaries in the microstructure. After cooling
again, the gear is reheated to 1500F (816C) to
austenitize the carburized steel. When the gear is
quenched in oil, much of the austenite transforms
into hard and brittle martensite. A deep freeze cycle
may be used to complete the transformation and
minimize retained austenite. Finally the gear is
tempered at 300-360F (149-182OC) to reduce
brittleness and microcracking.
The carburizing and hardening cycles produce a
very hard, mattensitic layer on the surface with a
less hard, tough core. In addition, the carburiiing
and quenching processes cause high residual
compressive stresses in the surface for increased
material strength and resistance to fatigue failure. Carburizing process control

Gas carburizing atmospheres can be in equilibrium
with a wide variety of carbon content. Proper surface
carbon content must be maintained to achieve the
required properties after heat treatment. Surface
carbon content should be maintained between 0.6%
carbon and 1.0% carbon. Less than 0.6% surface
carbon will lead to low surface hardness which may
not meet final requirements. Surface carbon contents over 1 .O% may lead to excessive retained austenite and/or grain boundary carbide networks. With
higher surface carbon contents it may be necessary
to sub-zero cool to transform retained austenite to

In a typical process for alloy steel, gas carburiiing

diffuses carbon into steel from a hydrocarbon gas atmosphere while the gear is heated to 1660F
(899OC)or higher in a furnace. Carbon diision into
the steel is affected by (1) timeof exposure, (2) temperature in the carburizing atmosphere, and (3) percentage of carbon in the atmosphere. The higher

The gas carburizing atmosphere is controlled to

obtain the desired carbon content in the surface of
the gear. It is diicult to measure carbon potential
directly. Cften one or more components of the gas
mixture, such as CO,, CH4, H20, or 02 are
measured and the other component (usually CO) is
assumed to be in proportion. However, the unmeasured component may vary due to variable gas
source, air leak or carbon build up. If CO, hydrogen

concentration of carbon near the surface provides

or other componentsvary,the reactionswill shii and

maximum local hardness.

make the carbon potential relationship inaccurate.

9.12.1 Carburizing


AGMA Qll-A94

Shim stock (a thin, flat material sample) can be used

to accurately assess the initial carbon potential of
the carbutizing gas. However, conditions may be
different later when the gears are exposed in the
furnace. Other carburizing methods
Vacuum carburizing and plasma carburizing produce results similar to gas carburiiing. Vacuum
carburizing is often faster because vacuum carbuiiing is carried out at higher temperatures [I 80&
1950F (982-l 066 C)], rather than the
1650-l 800F (899-982 C) typically used ip gas
carburizing. Plasma carburiiing occurs in a DC
electrically charged furnace with the gears acting as
the cathode. A plasma is formed around the gears
which enhances the absorption of carbon at the
carburizing surface of the gear.
9.12.2 Nitriding
Nitriding is an alternate case hardening process
often specified for gears when distortion would be
difficult to control if the gears were case carburized
and quenched. In nitriding, nitrogen is introduced
into the surface of the steel at relatively low
temperature [925-1050F (496-566 C)] from a
nitrogen containing atmosphere such as ammonia.
A hard case is produced by the formation of hard
nitride compounds in the surface making quenching
Special steels are needed that contain elements
such as aluminum or chromium to form hard nitrides
during treatment. The steel is nitrided in the hardened and tempered condition. The process is
controlled by adjusting the dissociation of the
ammonia. More often the Floe process is used which
is a double-stage process analogous to the boostdiffuse cycle in carburizing. In the first stage,
dissociation level is controlled at 1530% by using a
temperature range of 925- 975F( 496-524 C)
producing a white nitride layer which is diised in
the secpnd stage by increasing the dissociation to
80-85%. The high dissociation can be achieved by
increasing the temperature to 1025-l 050F
(552-566 C) and using an external dissociation.
The remaining white layer should be limited to a
maximum allowable depth of 0.0005 inch (0.013

mm). Some critical applications require removal of

the white layer.
Even with the two-stage process, nitriding is slow,
taking about ten times as long as carburizing to
produce a specified case. The nitride process
produces a very hard case with minimum distortion.
Volume increases during nitriding and causes favorable compressive stress to build up in the case.
However, nitrides tend to accumulate more at tooth
edges. ion nitriding
Ion nitriding or plasma nitriding is similar to plasma
carbutizing in that a plasma is formed around the
work during treatment.
9.12.3 Selective direct hardening
Selective direct hardening produces a hard case by
heating the surface layer above the austenitiiing
temperature and rapidly quenching, while leaving
the core in the original condition. A medium carbon
steel is used with the required carbon already in the
steel and heat treated to proper core hardness.
Since a large proportion of the part remains cool,
thus stabilizing the material, distortion is much less
than it would be if the entire part were heated. The
higher carbon containing material may make
machining more diicult than a carbuiiing grade of
Induction hardening is the one method of selective
directive hardening that may be suitable for some
production aerospace gears. However, the induction hardening process should be developed to
control residual stresses and annealing between the
case and core. Two other surface hardening
processes, laser heat treatment and electron beam
heat treatment, are also being developed. Induction hardening
Induction hardening is achieved by using an
alternating current in a work coil that surrounds the
part to be heated. An alternating magnetic field is
established that induces a potential in the part
causing a current to flow in the closed circuit.
Heating is produced by the resistance to the induced
current. The rate of heating depends on the strength
of the magnetic field. The depth of the field varies
inversely with the frequency of alternation. The

AGMA 91%A94

higher the frequency, the more shallow the heating

If a circular coil is used to heat a gear then the tips of
the gear are coupled closer to the coil and thus they
heat more, resulting in a deeper case depth at the
tooth tips.
After the heating is complete, the current is turned off
and the part is quenched by synchronized jets of a
quenching fluid, usually water-based. Laser heat treatment
Laser heat treatment is a surface-hardening process in which laser energy heats the surface above
the austenitiiing temperature. When the source of
energy is removed, the part self-quenches by
diffusion of the heat into the mass of the part. The
laser causes a steep temperature gradient due to
the extremely rapid heating rate. As the rate of heat
input increases, the depth of hardening is reduced,
since the temperature gradient becomes steeper
and the surface temperature must be limited to avoid
To spread the laser energy over required coverage
area, the beam is usually defocused. Alternatively
oscillating optics or integration optics with a faceted
mirror may be used.
Laser beam penetration is controlled by power level
and rate of beam traverse. Penetration increases
with increased power and decreases with increased
beam traverse rate. Electron beam heat treating
This method is similar in principle to laser heat treating, except that heating is achieved by an accelerated stream of electrons instead of a light or infrared
beam. When the electron beam is turned off, the part
self-quenches. The electron beam heat treat process occurs in a vacuum environment. This requirement introduces some complications into the f&king. The electron beam is manipulated by magnetic
9.12.4 Residual stress patterns
One advantage of case carburized parts is that when
the treatment is properly carried out it produces
compressive stress at the surface. Compressive


stresses help counteract tensile stresses produced

during tooth loading and thus increase the expected
life. A part that has been carburized is heated above
the austenitizing temperature and then quenched.
The surface cools faster than the center of the
section because heat is abstracted from the surface
by the quenching media. The net result is that
transformation of austenite to martensite starts at
the case/core interface with an expansion as
martensite is formed. The case is the last material to
transform and expand to martensite, causing
compressive stresses because the core has already
transformed and restrains the case.
The conditions are different in selective hardening,
but the results are similar. Energy is transmitted
quickly into the surface resulting in a surface layer
heated above the austenitizing temperature. This
layer will later become the hardened case. When the
energy is turned off, rapid cooling progresses and
again the case is the last to transform and the
restraint induces residual compressive stresses as
the surface expands during transformation from
austenite to martensite.
9.12.5 Dimensional problems caused by heat
Heat treatment tends to cause more quality problems than any other manufacturing step. Heat
treatment causes dimensional changes due to
volume change resulting from phase transformation. Distortion occurs from a combination of aeometric factors and uncontrolled stress relief. These
two factors acting together often cause unpredictable results. Variables that contribute towards the
dimensional changes include:
- Variations in material composition;
- Residual stress differences;
- Size of part (within tolerance range) before
heat treatment;
- Surface condition;
- Cart&zing heating cycle;
- Carburizing atmosphere control;
- Depth of case;
- Quenching parameters;
- Quenching die dimensions;
- Post heat treatment.

AGMA 9ll-A94

Gear manufacturers try to bring the component size

under control in the finish grinding stage. If excess
material is left on the part prior to heat treatment,
there will be enough stock to enable the size to be
brought under control. However, if too much is taken
off, the most effective portions of the carburiied (or
nitrided) case are removed. Figure 17 shows
uniform material being removed from a tooth after
heat treatment. In the example shown in figure 18,
the tooth has distorted to the right. To correct the
profile, excess stock has to be ground from the right
side of the tooth. This has several serious
Size before
Heat Treatment

Figure 17 - Schematic of material ground from
a gear tooth

Heat Treatment

First, there is lackof uniformity in case depth leading

to uneven residual stress distribution. Second (and
worse) is that the gear appears satisfactory in a
nondestructive inspection, even though the performance of the gear will be less than optimum. Third, a
considerable thickness of material has to be
removed during grinding, increasing the probability
of grinding bums. Some problems that are blamed
on grinding can in reality be traced back to heat
treatment. Thus the effects of heat treatment have to
be considered before and after the process in both
the soft machining and hard finishing stages.
Selective direct hardening processes minimize
distortion and associated rework problems.
9.13 Gear through hardening
Some aerospace gears that are not critically loaded
may be heat treated by through hardening
processes to obtain the required mechanical
properties for the gear application. Typical material
hardnesses range from 32 HRC to 54 HRC with
allowable gear tooth bending and contact stresses
that are lower than those for surface hardened
The process includes annealing, normalizing,
quenching and tempering. For annealing, ferrous
alloys are heated to 1475-l 650F (802-899%) and
furnace cooled below 600F (316%). In normalizing, ferrous alloys are heated to 1600-l 800F
(871-982C) and cooled in air. Annealing and
normalizing are used as homogenizing pm-heat
treatment processes to reduce metallurgical nonuniformities.
For the quench and temper process, ferrous alloys
are heated to 1475-16OOOF(802-871 C) to form
austenite, followed by rapid quenching which
causes the gear to become harder and stronger by
formation of martensite. The gear is then tempered
to a specific temperature below 1275OF (691 C) to
achieve the desired hardness and strength with
improved ductility and toughness.

Figure 18 - Schematic of material ground from

a distorted gear tooth


AGMA 91%A94

IO Surface treatment
This information covers a variety of currently utilized
surface treatments, generally applied after hardening, that are used to enhance the durability of aerospace gears with respect to their resistance to metal
fatigue, wear, and environmental corrosion.

10.1 Introduction
Post hardening surface treatments are usually
employed on gearing to accomplish one or more of
the following:
- Raise the bending fatigue strength at the tooth
root fillet radii, as well as in the rim, web, hub and
integral shafts;


.g 80

- Increase the contact fatigue properties along

the tooth flanks near the pitch line;
- Improve the resistance of gears, bearings, and
other mating surfaces to adhesive wear
phenomenasuch asfretting, scuffing and galling;
- Improve the resistance to abrasive wear, such
as scuffing (scoring) from ingested abrasive and
wear debris;
- Improve lubricity, both for break-in and continuous operation;
- Increase resistance to general and to localized
corrosion (pitting, stress corrosion, etc.) during
storage, from aggressive service environments,
or due to contaminated lubricants.




Peak residual stress, ksi

Figure 19 - Fatigue strength in ground AISI

4340 (50 HRC) [ll]
Tensile residual stresses may be present at the
surface of gears after heat treatment, particularly in
high strength, through hardened steels tempered at
temperatures which provide only moderate relief of
quenching stresses. Other processes which may
also introduce residual tensile stresses in gears
include grinding, abusive machining, hard plating
(such as nickel and chromium) and hard thermal
Some of the non-traditional
spray coatings.
machining processes such as electro-discharge
machining (EDM) and electrochemical machining
(ECM) may also damage surfaces through the
formulation of a brittle recast layer by EDM or the
introduction of pitting and/or shallow intergranular
attack by ECM.

Tooth bending fatigue is of primary importance

since fatigue fractures of gear teeth can result in
catastrophic destruction of a gear train. Contact
fatigue pitting, wear, and corrosion, however, can
also lead to tooth fatigue fractures from a reduction
in load carrying area and/or through the formulation
of local stress raisers. Since most geartooth fatigue
cracking initiates at or very near the surface, fatigue
properties are highly sensitive to surface characteristics such as roughness, and to residual surface
stresses introduced during the manufacturing process. As an example shown in figure 19, surface
residual tensile stresses can lower fatigue strength,

On the other hand, surface hardening heat treatments such as carburizing and nitriding introduce
beneficial residual compressive stresses in gears.
Some of their benefits may be offset, however, by
the presence of shallow, partial decarburization and
intergranular oxidation in unground roots of some

while fatigue strength is raised by the presence of

carburized gears, and the presence of a brittle white

surface residual compressive stresses.

nitride layer in some nitrided gears. Honing, polish-


AGMA 911-A94

ing, burnishing and other superfinishing techniques are often eff e&e in removing fatigue lowering surface layers and defects, as well as in reducing surface roughness.
Another effective process for improving fatigue
properties by inducing high surface residual
compressive stresses in gears is controlled shot
The process is utilized for both
carburized case hardened and through hardened
aerospace gears. Controlled shot peening should
not be confused with shot blasting, grit blasting,
or abras.ive blasting. The latter processes are
employed for cleaning or abrading surfaces and do
not produce consistent residual surface compressave stress profiles or predictable increases in gear
fatigue properties.
10.2 Shot peening
Controlled shot peening is a surface cold working
process in which hard, spherical shaped media
(steel, ceramic or glass) are propelled at relatively
high velocity and at a nearly normal incidence angle
against a workpiece. Its purpose is to promote surface strain hardening and to induce predictable surface and near surface residual compressive
Each particle of round shot striking the surface acts
like a tiny peening hammer, producing a small
indentation or dimple. The surface fibers are
stretched (yielded in tension) by the dimple formation, which also forms a sub-surface hemisphere of
strained metal below the dimple. Overlapping
dimples develop an even layer of plastically deformed surface fibers and a sut+sutface zone of
strained material. If unrestrained by the core, the
surface of the workpiece would elongate under
these deformation induced strains. The greater
mass of unaffected metal in the core, however,
restricts this expansion, producing high magnitude
residual compressive stresses in the surface and
near surface layers.

forces between surface and core are balanced.

Since the offsetting tensile stresses act over a significantly greater cross-section than the compressive stresses, the tensile stresses are generally of
low magnitude. See figure 20. In shot ieening thin
sections, the depth of peening is controlled so as to
keep the core tensile stresses at moderate values
and also to prevent distortion of the workpiece.
+1 00
4 wnax
- 6$
- 12
SS is surface stress
CS is compressive stress
TS is tensile stress
Figure 20 - Example of residual stress profile
created by shot peening [12]
The maximum residual compressive stresses from
shot peening are at least as great as half the yield
strength of the workpiece material, providing that
the media used for peening has a minimum
hardness at least as hard as the workpiece. For
example, the use of regular hardness steel shot
(HRC 45-52) will not induce the same magnitude of
compressive stress in a carburized gear as special
hardness shot (HRC 55-62). This is illustrated in
figures 21 and 22.

The maximum compressive stress is generally located just below the surface and decreases with increasing depth. The depth at which the residual
compressive stress becomes zero is usually referred to as the effective depth of peening. Beyond

The stress profiles in these and subsequent figures

were determined by standard X-ray diffraction
techniques for measuring residual stress (SAE
J784a). Subsurface measurements are made after
away surface layers and are
corrected for subsurface stress gradients [20].
Since the use of the harder shot increases surface
roughness, it may be necessary to final finish after

this point, the compressive stresses are offset by

peening when surface finish is critical. Lapping and



residual tensile stresses so that the

may be used if the operation

does not


AGMA 911-A94

remove more than 10% of the compressive stress

Regular hardness shot without post peen finishing
may also be used; however, the maximum compressive stresses will be lower than that obtainable
with the harder shot.



- -500
- -1000


Depth in inches


Figure 21 - Peening 1045 steel at 46 HRC with

330 shot [13]

Depth in inches

Figure 22 - Peening 1046 steel at 62 HRC with

330 shot [13]
10.2.1 Crack arrest due to shot peening
Metal fatigue cracking takes place in two distinct
phases. The initiation phase encompasses the
development and early growth of a small crack,
almost always at a free surface. The propagation
phase is that portion of the total life during which the
crack grows to the point of failure. It is often difficult
to define the transition from initiation to propagation.
In assessing the effect of residual compressive
stresses in increasing fatigue strength and life, it is
usually simpler to consider fatigue from the standpoint of fracture mechanics theory. This assumes

that minute cracks or crack-like flaws are always

present in engineering structures and that fatigue


life is primarily dependent on the propagation or

non-propagation of these cracks, which in turn
depends on the stress conditions at the crack tip.
The theory of crack arrest due to the residual compressive stresses induced by shot peening is based
on the following:
- A crack will not propagate unless a tensile
stress forces it open near the crack tip;
- The crack tip will not open as long as a
compressive force acts upon it.
Some important points to be considered with
respect to shot peening are:
-Gears should always be cleaned and inspected
thoroughly for cracks before shot peening. If
cracksarefound, the cause should be thoroughly
investigated and corrected. If allowed, the parts
should then be repaired by blending or re-machining. Shot peening tends to obscure cracks
and should not be used as a method for repairing
-Shot peening is effective as long as the residual
compressive stresses do not fade out or relax
due to exposure to high temperatures (generally
500F max. for most steels) or to over-stressing
(applied stresses in excess of the yield strength);
- Shot peening effectiveness is greatest when
to an area of a part where there is an
applied stress gradient, such as in bending or
torsion, or in the presence of geometric notches,
such as fillet radii;
-Shot peening is more effective in the high cycle
fatigue (HCF) domain; however, it is also beneficial in the low cycle fatigue (LCF) region up to the
point where the applied stresses correspond to
the net section yield strength.
10.22 Effect of shot peening on tooth bending
The tooth root fillet radii in aerospace gears are
generally subject to relatively high cyclic bending
stresses. Carburized case hardened gears, as well
as through hardened gears, are frequently shot
peened in the root and root fillet area to introduce a
high magnitude compressive stress, as illustrated in
figure 23. Significant increases in bending fatigue
strength can be realized when shot peening

carburiied and hardened gears with unground

roots, as well as in gears with ground rcotswhen the
peening is performedafter grinding. See figure 24.

AGMA 911-A94

Non-cased areas in carburized gears, such as the

fillet radii at rims, webs, hubsand shafts, also exhibit
improved fatigue properties after shot peening. See
figures 25 and 26.
NOTE - As bending stresses approach the yield
strength of the material, the effectiveness of shot
peening is reduced due to plastic deformation which
causes fading of the residual stresses.

T 5

Spiral bevel gear, 3.625 DP

Material, 8620H bar stock, heat treated:
Carburizsd & hardened to 61 HRC
Load bansmitt& 135 HP @ 5000 RPM
Shot peeningspecifications:
Shot size
170 H
010 -014A






Cycles to failure
Figure 24 - Increase in fatigue resistance of
spiral bevel gear 1151
Figure 23 - Stress profile of carburized gear
tooth root, ground and then shot peened with
special hardness shot [14]




zz 280

Shot peened - 43 ksi (300 N/mm2)


5x106 107
5x105 106

\ .



1 1

Cycles tti failure

Figure 25 - Fatigue tests on

rear axle shafts [16]

Figure 26 - Fatigue tests on

notched shafts [17]


AGMA 911-A94

109.3 Effect of shot peening on tooth contact


investigator has also indicated that this lower friction

can also reduce gear noise.

Contact fatigue pitting and spalling in gear teeth

generally initiates at the dedendum surface of the
tooth flank, just below the pitch line. Surface initiation results from the combined effect of subsurface
shear stresses from rolling and tensile frictional
traction stresses from negative sliding (rolling and
sliding in opposite directions). Testing of peened
and unpeened carburized and hardened gears has
indicated that shot peening of the tooth flanks increases contact fatigue life by a factor of about 1.6.
See figure 27.

The modem delamination theory of wear suggests

that all adhesive-wear processes are related to
fatigue and that the residual compressive surface
stresses from shot peening, as a result, may also
increase resistance to wear. This has not as yet
been confirmed in any rigorous test program.
10.2.5 Effect of shot peening on corrosion
In gear steels, shot peening has no effect on either
general corrosion or pitting corrosion.
peening, however, is effective in retarding the
initiation of fatigue cracking from corrosion pits until
the depth of the pits exceeds the effective
compressive depth of the peening.

10.2.4 Effect of shot peening on wear

High hardness gear steels exhibit only low to
moderate strain hardening when shot peened. The
increase in surface hardness due to shot peening,
therefore, has little effect on increasing resistance
to adhesive or abrasive wear. The tiny indentations
on the gear flanks produced by peening, however,
act as very small oil reservoirs which can help to
promote better lubrication, thereby reducing fretting
and scuffing (scoring), as well as lowering operating
temperatures by reducing friction. At least one

Several of the ultra-high strength, through hardened gear steels are susceptible to stress corrosion
cracking (SCC) when exposed to a moist chloride
atmosphere. The residual surface compressive
stresses from shot peening, often in combination
with a sacrificial coating (such as cadmium, aluminum and zinc) may prevent or retard SCC in these
steels as well as fatigue cracks initiated by SCC.

1 -8











Number of cycles xl O6
Standard gears
Comparison of surface (pitting) fatigue lives of
standard ground and shot-peened carburized
and hardened CVM AISI 9310 steel spur gears:



Number of cycles xl O6


Speed - 10 000 RPM

Lubricant - synthetic paraffinic oil
Gear temperature - 77C (170F)

Figure 27 - Fatigue life comparison 1181




AGMA 91%A94

10.2.6 Shot peening process controls

There is presently no non-destructive production
method to determine the proper shot peening of a
gear. Strict control of the shot peening process is,
therefore, essential to ensure repeatability and
uniformity on a part-to-part and lot-to-lot basic,
while conforming to applicable specifications. The
shot material, size, shape and hardness, as well as
velocity and impact angle, must be rigidly controlled
to provide consistency in peening results. In order
to effectively control the shot peening process, the
following parameters must be addressed:
- Media control
- Intensity control;
- Coverage control;
- Equipment control.
Details can be found in shot peening specifications
such as MIL-S-1316X, AMS2430, AMS2431 and
AMS2432, as well as the SAE Manual of Shot Peening, J808a. The following, therefore, is meant as
only a brief summary. Media control
Media typically used for shot peening are small
spheres of cast steel, conditioned cut wire (both
carbon and stainless steel), ceramic and glass materials. Peening media must be uniform in size and
essentially spherical in shape with no sharp edges
or broken particles. Broken or sharp edge particles
can be potentially damaging to the part surface. intensity control
Calibration of the impact energy orpeening intensity
of the shot stream is essential to controlled shot
peening. The energy of the shot stream is a function

The height of this curvature when measured in a

standard Almen gauge is called arc height. There
are three standard Almen strips currently in use: the
A strip, 0.051 in (1.3 mm) thick for intermediate
intensities; the C strip, 0.094 in (2.4 mm) thick for
high intensities; and the Nstrip, 0.031 in (0.79 mm)
thick for low intensities. Three strip thicknesses are
required since the useable range of curvature on the
Almen strips is 0.004 to 0.024 inches (0.1-0.6 mm).
A comparison of the intensities of the A strip for the
C strip and the N strip, as indicated by arc heights, is
also shown. See figure 28.
Intensity designations must include both the arc
height and type of strip used. Substitution between
strips is not permitted. An Almen arc height is not
properly termed intensity unless saturation is
achieved. This is done by developing an intensity
saturation curve. Saturation is defined as the
earliest point on the curve where doubling the
exposure time produces no more than a 10%
increase in arc height. Most important to the user is
the fact that the depth of the compressive layer is
proportional to the Almen intensity.



of the media size, material, hardness, velocity and

impingement angle. In order to specify, measure
and calibrate intensity, a method developed by J. 0.
Almen utilizing SAE 1070 spring steel specimens
called Almen strips is still in use. The unpeened
strip is fastened to a steel block and exposed to a
steam of peening shot for a given period of time.
Upon removal from the block, the residual compressive stress and surface plastic deformation produced by the peening impacts cause the Almen strip
to curve, convex on the peened surface.



0.001 inch




- 10







Almen intensity of C strip, mm

Figure 28 - Correlation of Aimen intensities as indicated by arc heights of A, C, and N strips
peened under identical blast and exposure conditions.


AGMA 91%A94 Coverage control

Coverage is defined as the extent, in percent, of
uniform and complete dimpling or obliteration of the
original workpiece surface. Inspection of percent
coverage can be accomplished by using a ten
power magnifying lens or through the use of an
approved fluorescent liquid tracer system. It is
extremely difficult to visually determine coverage in
hardened gears with a ten power visual examination. For determining coverage on gears, the liquid
fluorescent tracer system is widely used. The
coating, which is applied prior to peening, iS
removed during peening at a rate proportional to the
percent of shot peening coverage. Examination
under black light (UV) provides a practical and
superior method of verifying the 100% minimum
coverage required when shot peening gears for
fatigue resistance.
102.6.4 Equipment control

When shot peening is selected for man-flight vehicle geared components, for geared components
where it is used as part of the design strength, or for
geared components which are considered critical to
system success, the following are suggested:
- Shot peening by computer monitored/controlled equipment per an applicable, approved
specification such as AMS2432, including the
use of liquid tracer systems to verify coverage;
- Prior approval of a strength and life analysis
which justifies the approach. This includes life
certification/qualification testing on hardware
conforming to the production configuration;
-The establishment and validation of an inspection interval that takes into account the potential
for in service degradation of peening benefits;
- Any repair/rework peening required at prescribed inspection intervals to attain the required
fatigue life.

The machines used for shot peening provide means

for propelling shot by air pressure or centrifugal
force against the workpiece, as well as mechanical
systems for moving the work through the shot
stream or moving the shot stream through the work
by translation and/or rotation. The equipment also
continuously removes broken or defective shot so
that it is not used for peening. Many modem shot
peening machines are computer controlled. Typical
peening parameters which are monitored, controlled and documented are:

10.2.8 Guidelines for media and intensity selection

- Air pressure of each nozzle or wheel speed of

each wheel;

- Requirements for peening internal surfaces

and/or intersecting holes.

- Shot flow of each nozzle or wheel;

- Part rotation or translation rates;
- Nozzle reciprocation rates, distances and run
- Total cycle time.

The proper selection of media type, size, hardness

and intensities for peening of aerospace gearing is
dependent on numerous variable including:
- Minimum fillet radii in gear;
- Thickness of peened section of gear and/or
effective depth of carburized case, if any;
- Hardness of gear in areas to be peened;
- Surface finish requirements;

The tables of guidelines found in numerous specifications, such as MIL-S-13165C, are valuable statting points for selection.
NOTE - Magnetic particle, penetrant, ultrasonic or
otherflaw andcrack detectionmethodsshouldbe performedprior to shot peening.

102.7 Shot peening design considerations

10.3 Surface coatings

When the gear user does not wish to take advantage of the fatigue strength increases in the design
calculations, but wishes to use the process to overcome many of the residual surface tensile stresses
or similar problems from previous manufacturing

Table 16 lists of some of the most widely used coatings for aerospace gears. The table is divided into
three broad categories of coating application: antifretting barriers, corrosion resistance, and build-up

processes, shot peening controlled per AMS2430

Under the anti-fretting barrier list are coatings used

or the equivalent is generally suitable.

to reduce the effects of fretting at joints. Aluminum


AGMA Qll-A94

bronze econal has been successfully used between

mounting surfaces of gears which are bolted to
flanges. Silver and copper plate have also been
used on gear bolted connections, but are more
normally used as coatings for splines. Tungsten
carbide has had application to cone seat connections between helicopter rotor heads and main rotor
shafts under high loads and stresses.
Steel parts exposed to the atmosphere have used
baked resin (or other forms of paint) in non-working
areas for corrosion resistance. Cadmium plate has
been used on steel parts exposed to the
atmosphere with brush cadmium being used on
high strength steel (over 150 ksi) and conventional
cadmium on lower strength components. Thin
dense chrome has been used on grease lubricated
bearings exposed to the atmosphere. There is
disagreement over the benefits vs. cost of black
oxide or phosphate coatings on steel parts.
The third broad category of coatings used in
aerospace applications is in the area of repair of
surfaces that are under the minimum material
condition. Sulfamate nickel plate has been used as
a repair for bearing journals that are up to 0.010 inch
undersized, in carburized and uncarburized
conditions. Since there may be a reduction in
fatigue strength from sulfamate nickel plating,
repairs should be made in areas of low stress only.
Electroless nickel and brush nickel plate have been
used for very thin coatings where precision is
required. The parts need not be ground after plating
as the thickness is uniform and the plating takes the
form and finish of the base material.

fields. The ion implantation process is primarily

used for two reasons:
- Improving surface hardness, wear resistance
against adhesive, abrasive and scuffing conditions;
- For improving corrosion



In the ion implantation process, energetic ions

impinge on the surface of interest. The ions
penetrate into a substrate material, modifying the
surface by changing crystallinity and chemical
structure of the material.
Ion implantation has proven to be an effective and
technically attractive approach for changing the
surface properties of high value added and high
precision components of various criiical systems.
Ion implantation of gears was introduced and has
been used in the last ten years for a variety of
reasons, primarily for improving the mechanical
performance in a wide range of applications and
improving corrosion resistance in helicopter
transmissions and aircraft engines.
10.42 Improved mechanical performance of the
Ion implantation has been used on a variety of gears
for improving surface hardness, providing lubrication and minimizing scuffing wear of gears. The
most common material studied in this field is AISI
9310 steel, a commonly used material for gear

10.4.1 Introduction

Ion implantation is also used extensively for

treatment of titanium alloy gears. In aerospace
applications, for weight saving purposes, titanium
gears are occasionally used. Ion implantation of ion
species, such as nitrogen and/or carbon, into
titanium alloys will induce hard phase precipitates of
titanium nitride and/or titanium carbide. The
increased surface hardness of the gears and lower
coefficient of friction induced by ion implantation
has proven very effective in reducing scuffing wear
and lowering wear of titanium components.

Ion implantation isan effective processforchanging

the surface properties of materials without
adversely affecting the bulk properties.
implantation technology is used in a variety of
applications in aerospace, automotive, cutting tool,
biomaterials, metal stamping and metal piercing

In hydraulic fuel applications, ion implantation has

been used to modify the surface properties of the
face of the gear. In this case, the main concern is
excessively high friction, adhesive wear and particle
debris generation. ion implantation of titanium and
carbon into commonly used steels used in fuel

For any coating that is applied in areas that are

clamped or otherwise in contact, consideration
must be given to the consequence of loss of coating
thickness, loss of preload on bolted joints, loss of fits
for press fitted parts, etc.
10.4 Ion implantation of gears


AGMA 911-A94

pumps has proven very effective in minimizing

friction and improving resistance of the surface
against wear.
10.4.3 Ion implantation for corrosion resistance
ion implantation has been used extensively to
modify corrosion resistance of steel gears. The
most common recipes include ion species such as
chromium, molybdenum, tantalum and titanium
followed by carbon ion implantation. Depending on
the nature of corrosive attack, one or the other of
these ion species have been studied in laboratory

testing, as well as field appiications. Implantation of

ion species such as chromium and molybdenum
have significantly improved pitting corrosion and
aqueous corrosion in salt bath applications.
10.4.4 Conclusion
High precision gears appear to be ideally suited for
the ion implantation process. The available data
supports the use of the technology on a broader
range of applications requiring superior corrosion
and wear resistance.

Table 16 - Surface coatings used in aerospace gear units




Aluminum bronze econal
Silver plate
Sulfamate nickel plate
Tungsten carbide
Detonation thermal spray
Tungsten carbide plasma
Corrosion resistance

Fretting barrier (plasma spray)

Fretting barrier
Fretting barrier (SplinesIJoints)
Limited fretting barrier
Hard, wear resistant, fretting barrier


Hard, wear resistant, fretting barrier


Baked resin
Black oxide
Brush cadmium plate

Corrosion resistance
Corrosion resistance (on shelf)
Corrosion resistance.
Used on high strength parts
Corrosion resistance for
parts outside gearbox
Touch up for Black Oxide
Corrosion resistance
Corrosion resistance &
break in surface for gears
Corrosion resistance, hard wear surface


Thin build up repair (0.0002 inches thick).

Precision plate with no grinding.
Thin build up repair (0.0002 inches thick)
No machining after plating.
Uniform plating thickness even in grooves.
Loss of fatigue strength.
Build up repair (0.005 inches thick)


Cadmium plate
Gun blue
Ion vapor deposition
Phosphate coat
Thin dense chrome


Build-uo reoair
Brush nickel plate
Electroless nickel plate

Sulfamate nickel plate




AGMA 911-A94

11 Manufacturing


11.I Introduction
The performance of an aerospace transmission is
directly related to the manufacturing process used
to fabricate the gears. Due to the unique characteristics of the sliding mesh, close control of the gear
tooth geometry is paramount to yield a successful
application. The metallurgical and geometric quality must be controlled to meet the drawing requirement. In addition, surface finish, including surface
roughness and waviness, is important to reduce
surface distress. The Engineering drawing requirements should dictate the manufacturing source and
process used to fabricate the gears. The selection
process for the manufacturer should include the
approval of Engineering, Manufacturing Engineering, and Quality Assurance. The manufacturer
should provide tangible evidence of their capability
to meet the drawing requirements.

11.2 Spur and helical gears

Manufacturing spur and helical gears require
unique machines to cut (generate) the gear teeth
prior to the carburizing and hardening process.
Generally, the parallel axis gear teeth are semi-finished, using a shaping or hobbing machine. The
shaping or hobbing cutters should be designed to
accommodate the appropriate stock on the gear
teeth to meet the needs to clean up the gear teeth
after hardening. Care should be taken not to leave
too much stockon the teeth, which would reduce the
effective case depth and hardness. The tooth form
(involute and lead) and tooth accuracy (pitch and
index variation) should be controlled within reasonable tolerances to assure proper stock allowance
for final grinding of the teeth. Sometimes, 200 percent of the drawing tolerance for the above requirements can be used to control the as-cut dimensions. The actual amount needs to be carefully
chosen based on the material and configuration.

11.2.1 Shaping gears

The shaping machine process can be used to generate internal and external gears, especially gears
which have a shoulder or other physical obstruction
limiting the path of thecutter. Shaping machinesare
unique in that they can generate most or all parallel
axis gearswhen properly tooled and fuctured. Shap
ing helical gears will require a special helical guide.

Special shaping machines are available to shape,

crown, and/or taper the spur or helical tooth. The
involute profjle is a function of the involute shaping
cutter and the generating machine motion.

11.2.2 Hobbing gears

Hobbing is recommended for semi-finishing
external spur and helical gears where the blank
configuration permits the clearance for the cutter
path. Hobbing is generally the most cost effective
means of generating spur and helical gears.
Hobbing machines can generate crown and/or
taper. Helical gear teeth are generated through a
differential gear box or by computer numerical
control (CNC)/electronic control.

112.3 Semkfinish tooth geometry

The semtiinish tooth geometry and accuracy
should be controlled by measuring the first part and
last part on a qualified measuring machine. Additional parts should be measured when a cutter is
changed or the machine setup is changed. Generally, gears are semi-finished by cutting on a hob or
shaping machine but may be ground from solid
(directly from a blank), or milled, forged, rolled, etc.

112.4 Heat treat considerations

Consideration should be given to heat treat
It may be required to adjust the
measurement over or between pins to accommodate dimensional change in the carburizing and
hardening cycle. A lead adjustment may also be
required to counteract lead variations introduced by
Generally, a pilot lot should be
heat treat.
processed through the heat treat cycle to confirm
the effects of hardening. The larger the gear
diameter and the longer the face width, the more
critical this process may be. For consideration of
microstructure, see clause 9 of this document.
Nitriding and induction hardening of some aircraft
gears are also used. These processes reduce
much of the distortion of the gear teeth and the gear
blank experienced during the carburizing and
hardening process.
The nitriding process produces a thin, very hard,
brittle layer in some materials that may cause damage in certain applications. Some manufacturers allow this material to wear off during break-in. Others
remove it by either light grinding or by electrochemical machining (ECM). This non-abrasive


AGMA 91%A94

process removes the white layer with chemicals that

slowly dissolve the surface. A non-destructive
chemical check can be performed to determine if
the white layer has been sufficiently removed.
112.5 Gear blank control
Gear blank dimensional control before, during, and
after heat treat is important. The operation sheets
are used to control the manufacturing process,
including the dimensions and tolerances for the
surfaces used to support the work holding fixtures
and the surfaces used to contact the quench die
where applicable.
The gear tooth accuracy is directly related to the
control of the tooling points of the gear blank. In
addition, the contact points related to the quench die
will dictate the consistency of the gear teeth through
the hardening,process.
112.6 Grinding gear teeth. Finish grinding gear
teeth is important to provide accurate and consistent gears. Selecting the grinding machine may be
the most important decision to assure the gear teeth
will meet the engineering drawing requirements for
geometry and accuracy. Every gear is unique and
one machine may be able to perform better than
others. The following are some suggestions in selecting a gear grinding machine:
- A shoukier or physical obstruction adjacent to
the gearteeth restrictstheselection of a machine
type. Form grinding permits the use of small
wheels in this case;
- Gears which require the root fillet radius to be
ground restrict the use of some gear grinding machines;
- Most aerospace gears have geometry and accuracy tolerances requiring gear grinding machines to be maintained to excellent condition. lt
is not unusual for gear designs to require AGMA
Class No. 13 and 14. In order to meet AGMA
Class No. 14 successfully, consistently, and economically, it is recommended to use electronic
and CNC grinders, or other machines that have
demonstrated they can meet this requirement.
Temperature control of the grinding machine and
the ambient temperature is important to obtain
the specified accuracy;


- The selection of the grinding wheel and the

method of dressing the wheel are factors in controlling the involute form and the surface finish.
The wheel abrasive, grii size, hardness, bond,
coolant oil, wheel surface feed, and machine
feed rates are all factors in the performance.
Plated cubic boron nitride (CBN) grinding wheels
can be used successfully to grind precision gears
with good involute and surface finish.
11.2.7 Measurement and control
All precision gears should be measured and controlled based on engineer requirements. Special
requirements will include involute, lead, pitch, and
index variation and should be measured on a precision certified measuring machine. As the precision
and accuracy requirements of the gear increase, so
does the accuracy requirements for the measuring
system. A calibration method should be established
to certify the accuracy of the measuring machine.
11.2.8 Other measurements
In addition, the major and minor diameter (whole
depth), root fillet radius, and measurement over or
between pins and/or balls (tooth thickness) should
be measured and controlled. Surface finish, roughness, and waviness are important elements to be
measured and controlled. Catburized, hardened,
and ground gears should be evaluated for abusive
grinding (burning) by surface temper inspection.
112.9 Deburring
Deburring and radiusing the edges of gear teeth
shoukl be accomplished using care not to damage
the working surface of the gear teeth and use
caution not to abuse (overheat) the surface, which
may cause surface temper on the edges.
11.3 Bevel gears
This document will address spiral bevel, straight
bevel, and Zero1 bevel gears. It should be
understood that the spiral bevel gear tooth form is a
three dimensional curve, which is not traceable to
any known curve. The geometry of spiral bevel,
straight bevel, and Zero1bevel gears is traceable to
the machine that is used to manufacture the tooth
form. Most precision aerospace bevel gears are
ground using modified roll. Wiih the advent of the
CNC controlled spiral bevel gear grinding machines, which have the abilii to grind other tooth
fans, the aerospace industry will continue to use

AGMA 911-A94

the modified roll system for most spiral bevel gears

for some time. Straight bevel gears are usually not
ground because there are limited machines
available to grind straight bevel gear teeth.
Note: The majority of aerospace bevel gears are designed and fabricated using the Gleason system.

11.3.1 The source for data

The origin of the machine settings used on the
cutting and grinding machines are developed by
using the many computer programs available to the
industry. Initially, the Engineer will design the gear
mesh based on the requirements of the application.
The power, velocity, tooth load, shaft angle, ratio,
temperature, the mechanical application, and many
other characteristics are considered. The following
is some of the basic data used by the manufacturing
engineer who will develop the machine settings
necessary to meet the engineering design:

Number of Teeth;
Pressure Angle;
Diametral Piich;
Spiral Angle;
Face Width;
Shaft Angle;
Whole Depth;
Working Depth;
Root Angle;
Root Fillet Radius;
Addendum Angle.
Develop flank form geometry

Once the basic gear requirement is established by

Design Engineering, the Gear Manufacturing
Engineer develops the toothflankform geometry by
using a computer system. This is accomplished by
conducting the following computer analysis:
a) First, a dimension sheet which includes the
basic dimensions of the bevel gear and pinion
and also includes some of the strength and
durability factors for the design must be established based on the Engineering requirements;
b) Second, a tooth contact analysis is developed.
The computer has a working model of the
generating motion, including the kinematics of
the meshing of the gear and pinion teeth. In
principle, this program develops the tooth flank
form geometry on the computer in the same
mannerasthe machine operatordoes except it is

instant, and the Engineer has a more scientific

approach. The Engineer has the opportunity to
change any one or all of the eighteen machine
settings, which control(s) the gear tooth geometry of the gear, and the eighteen machine settings
which control the pinion. Conventionally, the
gear is cut and ground spread blade (grinding or
cutting both sides of the tooth at the same time),
but the Engineer has the option to produce the
gear single side in the same manner as the
pinion. The tooth contact analysis will display the
lines of contact and the path of contact at the toe,
heel, and mean of thetooth. In addition, the tooth
contact analysis displays the motion curves
(transmission error) at the toe, heel, and mean of
the tooth. During the development of the tooth
contact pattern required by the development of
the tooth contact analysis, the Manufacturing
Engineer is evaluating the following:
i. Lengthwise curvature;
ii. Tooth bias;
iii. Tooth profile curvature;
iv. The effect of the blade angle (longer at
the dedendum and shorter at the addendum, etc.);
v. Length of the toe versus the length of the
vi. The path of contact;
vii. The profile width at the toe versus the
profile width at the heel.
11.3.3 Engineering skills
The development of the tooth contact analysis
requires either second or third order changes in
order to correct for an undesirable pattern (load
distribution). The Manufacturing Engineers level of
experience and understanding of the gear mesh
application is directly related to the level of success
in this work Generally, further development work is
conducted at the time the gears are generated (cut
and ground).
11.3.4 Loaded tooth contact analysis
The Manufacturing Engineer has an option to run
the loaded tooth contact analysis program. This
program is similar to the standard tooth contact
analysis but can display the line and path of contact

as well as motion curves based on seven diierent

loads. This program may further define required


AGMA 911-A94

changes to the flank form geometry to improve the

load distribution. This would require changes to the
input data for the tooth contact analysis.

computer system. It will also provide the machine

changes required to correct the errant tooth form.

11.3.5 Summary of machine settings

Manufacturing bevel gears requires a good manufacturing plan. The Manufacturing Engineer should
have a good technical background in bevel gear
manufacturing. The manufacturing plan should use
good logic for processing the gear, taking into consideration the following:

When the tooth contact analysis has been developed to the satisfaction of the gear Manufacturing
Engineer, the summary should be produced. This
data provides the machine settings for cutting and
grinding the gear and pinion. In addition to machine
settings, basic gear data is given as well as settings
for the Blank Checker, which measures the face and
back angle, root depth, and root angle. Proportional
changes are given for making second order
changes to the pinion.
11.3.6 Grinding sequence
Running the Grinding Sequence Program will provide information for sizing the grinding wheel, type
of grinding wheel, feeds and speeds, and other sequence data necessary to grind the gear and pinion.
11.3.7 Fillet details
The gear Manufacturing Engineer has the option of
running Tooth Profile and Fillet Details program.
This program will define the generated root fillet radius of the gear and pinion based on the generator
and grinding machine settings.
11.3.8 Current technology
Todays current technology permits defining the
working flanks of spiral bevel gears scientifically
using a three-dimensional, coordinatesystem. The
computer system will define the theoretical tooth
form established in the special analysis file located
in the mainframe computer, while developing the
tooth contact analysis and summary of machine
settings. The theoretical coordinate data established on the computer can be downloaded to the
coordinate measuring machine for use in measuring theflankform of the spiral bevel gear teeth. With
the software system, it is possible to get first and
second order corrections to the machine settings
used to produce the gear teeth. That is, a gear
produced on a cutting or grinding machine can be
measured on a coordinate measuring machine
using the computer program. It will define the

variance of the machined surface as compared to

the theoretical or measured data stored in the


11.3.9 Manufacturing bevel gears

a) The dimensional tolerances on the Engineering drawing;

b) The gear tooth geometry and accuracy;
c) The heat treat distortion and processing;
d) The finish processing, including deburring,
plating, and final inspection;
e) Each operation should be described clearly
and concisely using picture sheets when
11.3.10 Generating
Generating the gear teeth prior to carburizing and
hardening is important. The primary considerations
are as follows:
a) Controlling the flank form geometry to permit
minimum stock removal for finish hard grinding
the gear teeth;
b) Control the root depth and root fillet radius to
permit proper stock for finish grinding the gear
c) The tooth accuracy, pitch, and index variation
should be controlled to provide the requirements
for hard grinding the gear teeth;
d) The surface finish should be controlled to
assure that the cutter flats and scratches do not
endanger cleaning up the tooth surface while
finish grinding gear teeth.
115.11 Heat treat distortion
Consideration of heat treat distortion during all
phases of gear blank fabrication and generating the
bevel gear teeth must be maintained. Carburizing
and hardening gears, which causes the gear blank
to distort, has always proven to be difficult,
especially on gears with large diameters and thin
cross sections. The spiral angle of spiral bevel gear
teeth tends to unwind during heat treat. To control
the stock removal of the carburized gear teeth, the

AGMA 911-A94

face and back cone and crown diameter must be

maintained within close limits. Quench dies are
generally used on large gears by placing contact
pressure on the outside diameter, inside diameter,
and top face angles during the quench cycle to
reduce some of the distortion. Gears must be flat
and round after hardening.
Nitriding and induction hardening are also used for
some bevel gears. These processes reduce much
of the distortion of the gear teeth and the gear blank.
11.3.12 Control after hardening
Controlling the gear blanks after hardening is
accomplished by using a good manufacturing
process and tooling. The pitch line of the gear teeth
must be concentric to the mounting diameter and
mounting surface. This part of the manufacturing
process is critical in order to maintain stock removal
within limits of the case depth and surface hardness
on the gear teeth. The accuracy of the gear teeth
are traceable to the accuracy of the gear blank.
11.3.13 Grinding bevel gear teeth
Grinding the spiral bevel gear teeth is generally the
last machining operation and is among the most
important. Using the grinding machine that can
meet all the requirements of the engineering
drawing is the first and most critical step. Other
considerations are as follows:
a) Using the established summary and final machine settings for the gear grinder;
b) Controlling and recording the machine settings used to grind the gears;
c) Stock dividing the gear to the grinding wheel to
remove equal stock from the flanks and control
stock from thwoot of the gear tooth while maintaining the root depth requirement;
d) Use the grinding wheel, which maintains form
and produces the required surface finish without
tempering the gear tooth.
11.3.14 Measurement of bevel gears Tooth flank form
Measuring and controlling ground bevel gear teeth
is among the important steps in fabricating bevel
gears. The recommended method for measuring

bevel gear teeth flank form is using the coordinate


This method is the most scientific and

accurate system. An alternate method is using the

rolling method on a test machine for taking tooth
patterns. The tooth pattern record should include
vertical (V) and horizontal (H) measurements to the
toe and heel to control bias, and (V) only
measurements to the toe and heel to control the
pattern length (see AGMA 390.03a). Additional
V&H measurements should be taken to control the
tooth profile width. Other measurements
The root depth and root fillet radius must be
measured and controlled to meet the Engineering
drawing requirement. The surface finish, pitch and
index variation, and face and back angle of the gear
should also be measured and controlled
11.3.47 Deburring and radiusing
Deburring and radiusing gear teeth must be
accomplished, being cautious not to temper the
tooth edges after the gear teeth are ground.
Abusive grinding during the debut-ringand radiusing
can be detrimental to the strength of the gear tooth.
Surface temper inspection should be performed
after final machining, including deburring to assure
that the gear is free of temper. Automated deburring
and brushing are the preferred methods.
NOTE - One way to avoid surface temper is to adequately chamfer the gear teeth prior to hardening.

11.3.18 Final processing

The final processing of the gears, which includes
surface temper inspection, plating, shotpeening,
magnetic particle inspection, and coating must be
conducted according to approved operating
11A Stress relief treatment
Stress relief should be performed on all gears that
have been carburized and hardened as soon as
possible after grinding and should be performed
prior to any subsequent processing. 9310 steel
parts should be stress-relieved at 275 to 300 F
(135 to 150 C) for at least one hour (depending on
the cross section of the gear) followed by still air or
oven cooling to room temperature. Stress relief
treatment will minimize the possibility of residual
stress cracking. Since the possibility of such
cracking increases with time between grinding and
stress relief, it is recommended that stress relief be
performed within eight hours of grinding.

AGMA 91-l-A94

12 Gear inspection
12.1 General
This clause describes the recommended methods,
practices and controls used for the assurance of the
desired quality level of aerospace gears. The
clause is divided into two parts and discusses spur
and helical involute gears and bevel gears. The
dimensional and gear tooth element inspection of
aerospace gears is the same as for gears used in
other applications. However, the controls which
maintain the integrity, reliability and life are more
stringent and more supportive dccumentation and
records are required.
12.2 Spur and helical involute gears
The inspection methods and practices for the
inspection of spur and helical involute gears are
comprehensively discussed in ANSVAGMA
2000-A88, Gear Classification and inspection
Handbook, Tolerances and Measuring Methods for
Unassembled Spur and Helical Gears (including
Metric Equivalents). The following covers additional considerations for the inspection and control
of aerospace spur and helical involute gears.
122.1 Identification/traceability
Because the integrity and lie of aerospace gears is
so important, and in the event of a failure having to

retire other gears from service, it is recommended

that all gears be permanently identified with a lot
number or with an individual serial number. The lot
number or serial number should be traceable to a
raw material heat and melt number and be in
evidence throughout manufacture and be documented at significant manufacturing, inspection or
processing operations. Traceability with serial
numbers can significantly assist in the investigation
of any problem or failure.
12.29 Heat treat verification
The verification of final heat treat results is of the
utmost importance, particularly if the gears are
processed to produce a high hardness layer such as
by carburizing or nitriding. A metallurgical laboratory analysis of an actual part, part section, or
coupon (see figure 29 for coupon example) representing the gear teeth and processed with the lot of
gears is the recommended method of assuring the
results of the heat treating process. This analysis
would be traceable to the lot either by permanently
ident@ing the part with the heat treat lot number or
by documentation of part serial numbers. The
laboratory analysis would also be used to calculate
the minimum and maximum amount of stock that
can be removed to meet the specified finish depth of
the hardened layer.

Surface A, TYP, 2 PL

Drill thru 306 (0.187) diameter

Figure 29 - Heat treat coupon


AGMA 911-A94

122.3 Hardened layer finished depth

When required to assure that the finished depth of
the hardened layer meets specification, the recommended method is to monitor through process
control at the grinder by stockdividing and recording
the machine in-feed setting at the first contact with
the gear and the final machine setting. The stock
removal can then be easily calculated. This method
proves the entire process of the gear from the
calculated root and flank stock through cutting, heat
treat, quench and final gear grind.
129.4 Surface temper inspection
Inspection of high hardness ferrous gears,
particularly those with carburized surfaces, for
abusive grinding should be accomplished by
in-process surface temper inspections during the
grind operations. It is also recommended that all
carburized gears be resurface temper etched after
all machining and deburring operations are
complete. A discussion of surface temper may be
found in AGMA 230.01, Surface Temperlnspection
Process. The requirements for this process should
be mutually agreed upon by the manufacturer and
the purchaser because failure to detect abusive
grinding can have serious consequences.
122.5 Inspection

for surface or subsurface

Essential to assurance of the reliability, lie and

control of the cost of gears is the need to verify the
integrity of their material condition at various stages
of manufacture. If there is a concern that the raw
material, i.e. forgings or bar, may contain
undesirable characteristics such as laps, seams or
inclusions they should be non-destructively tested
prior to or immediately after final heat treatment.
The most commonly used tests are magnetic
particle, which detects surface or slightly
subsurface defects, or ultrasonic which will detect
all defects through the material.
If gears are shot peened, a non-destructive test of
the shot peened areas prior to peening will assure
that no defects open to the surface will be peened
An appropriate final non-destructive test should be
performed after all manufacturing, plating and
coating operations are completed to assure that no

manufacturing, plating or coating operation has

induced defects that would be detrimental to the use
of the gear.
12.2.6 Inspection of root fillet radii
Inspection of the root fillet radii should be accomplished by the use of a magnifying tracer instrument
which charts the radius form, or by the use of a cast
and optical comparator. To assure maximum tooth
strength, the fillet radii should meet specifications
for size and smoothness.
12.2.7 Verification

of dimensions

after shot

Precision gear configurations that are shot peened

should have close tolerance features such as
diameters, lead and runout m-inspected after shot
peen to assure that no undesirable dimensional
change has taken place. Refer to clause 10 for
other shot peen requirements.
12.2.8 Conformity inspections
Assurance that a first run gear has been
manufactured correctly to the design requirements
may be accomplished through a comprehensive
destructive test of a completed part by experienced
laboratory personnel. This activity can also be
conducted on a periodic basis through the
manufacturer/purchaser agreement to assure
process control.
12.3 Bevel gears
The inspection methods and practices for the
inspection of bevel gears are discussed in AGMA
390.03a, Gear Handbook - Gear Classification,
Materials and Measuring Methods for Bevel,
Hypoid, Fine Pitch Wormgearing and Racks only as
Unassembled Gears. The following covers additional considerations for the inspection and control
of aerospace bevel gears.
12.3.1 Master gears
To successfully maintain proper pattern position
and interchangeability during the production of
bevel gears, it is necessary to establish master
gears during preproduction development of the
gear sets. When the proper working pattern is
achieved at least three sets of mating gears should
be selected by evaluating the tooth contact patterns
on bevel gear test machines and designating them
as master gears at the following levels:


AGMA 91%A94

a) The top level master set is selected as the best

set to duplicate successfully tested gears;
b) The second level master set is usually
designated as the reference master set and is
periodically tested with the top level master mate
to assure that the correct contact pattern is
maintained for transfer to the third level master;
c) The third level master sets are used to test the
production gears for contact pattern position and
backlash and are normally called working masters or working control masters.
To preserve the patterns from the master gears for
wear comparisons at a later date, tape transfers of
the pattern positions are made using a colored
compound on the gear teeth during testing and
lifting the contact pattern from the tooth with a
transparent tape to a white hardboard card.

Wih the proper equipment and software, pitch,

index, spacing and the finished tooth surface tooth
form, as well as other characteristics, may be very
precisely measured. Measurement of the tooth
flank form consists of a digitized map of the working
surface of either flank that is compared to a
computer stored digitized master tooth.
This method of tooth flank form measuring is being
used successfully in the gear industry to maintain
process control during manufacture. Also, the
technology is available to allow automatic calculations from the digitized data to corrected machine
settings for optimization of tooth form.
12.3.6 Surface temper inspection
Refer to 12.2.4.
12.3.7 Inspection

for surface or subsurface

Refer to 12.25.

The top level masters should be well protected,

preserved and used only to test second level

12.3.8 Inspection of root fillet radii

Second level masters may require replacement

occasionally, dependent upon production volume
and the frequency of testing with third level masters.

12.3.9 Verification

Refer to 12.2.6.
of dimensions

after shot

Refer to 12.2.6.

Third level masters must be controlled closely

because they are used frequently to test the working
production gearing.

12.3.10 Conformity inspections

12.3.2 Identificatio~nkaceability

13 Rocket and space getiring

Raw Material Control and Traceability - Refer to


13.1 introduction

Refer to 12.2.8.

Verification of Stock Removal - Refer to 12.2.3.

However, the calculation of the stock removal may
be somewhat more complex due to the mechanics
of the bevel gear grinder.

Gears are in common use in space vehicles of all

types. As experience is accumulated in space, it
becomes increasingly clear that space imposes its
own set of rules for survival. As the distance of
operation from the earth increases, the environment becomes a dominant factor in the design
equation. The behavior of many materials changes
in the vacuum of space, and these changes must be
considered in the design of space mechanisms.

12.3.5 Coordinate measuring machine


Vehicles designed for operation in space fall into

three general categories:

12.3.3 Heat treat verification

Veriication of Heat Treat Results - Refer to 12.2.2.
123.4 Hardened layer finished depth

In addition to the measuring methods and practices

for bevel gears that are described in AGMA

a) Rockets - contain single use, short life span,

high power gearboxes or mechanisms. These
units are expended by a single flight;

machines may be used to inspect many features of

b) Reusable spacecraft and space planes -

bevel gears.

gearing and mechanisms must be suitable for


AGMA 911-A94

repeated reuse over a series of flights and

launches. Operation in a harsh environment
usually involves a short mission duration typically
measured in days or weeks;
c) Space Station and Satellites - geared
systems for these applications must survive their
operating environment for extended periods of
time (to more than ten years). Functional
reliability is paramount and there is no opportunity for preventive maintenance.
13.1 .I Rocket gearboxes
One of the first applications of gears for space was
in the liquid fueled rocket. The turbopump gearing
on these vehicles transmit the power to pump the
thousands of gallons of fuel consumed during the
first few minutes of flight.
All of the considerations for light weight power
gearing outlined in clause 4of this information sheet
apply to rocket gearing. The power density for
rocket gearing is extremely high since the mission
time is short and minimum weight is critical. In
qualification tests, these gears must function successfully for approximately 5000 seconds at full
13.1.2 Spacecraft and space planes
These craft embody the use of all the usual
mechanisms found on aircraft, such as control
surface actuators, bay door actuators, etc. In
addition, there are mission specific actuators used
in the deployment of satellites and other payloads
from the cargo bay. When the mission involves
manipulation of large bodies, the vehicle, such as a
space shuttle, is fitted with a robotic arm.
Mechanisms aboard space shuttles and planes are
typically designed for repeated usage, and therefore must withstand the vibration and shock of
repeated lift-offs and landings in addition to the
function they perform on a space mission. Since a
shuttle operates at an attitude less than 150 miles
above the earths surface, the environment includes
exposure to highly reactive atomic oxygen and
13.1.3 Space station and satellites
Although geared systems in this last category of

mission is extended and can be in excess of 10

years. As a consequence, the opportunity for
inspection and refurbishment does not exist, and
the mechanisms will be subject to degradation of
lubricant from evaporation and erosion, and extended wear on working surfaces.
13.1.4 Gear forms for space use
Many gear forms have found a place in space
mechanisms. These include extensive use of spur
gears, some bevel, helical and worm gearing, and
an occasional proprietary tooth form. Since light
weight structure is of paramount importance, gear
forms requiring high precision mountings are more
diiicult to apply successfully in the space mechanism. Similarly, forms that are relatively heavy for
the torque transmitted, or forms that exhibit a high
level of sliding at their contacting surfaces, are less
suitable for the space mechanism.
Motors which power space mechanisms are themselves small and usually have only modest torque
outputs. As a consequence, the efficiency of the
drive is of prime importance. In some drives, such
as those used on robotic arms, provision for
backdriiing in case of a collision or momentary
overload is a requirement. To achieve backdriieablilty in a gear train whose transmission ratio
may be in excess of 18OO:l points to the use of high
precision spur or helical gears.
Althoughoperating ina weightlessenvironment, the
payload mass which must be manipulated is often
large, and output torques of a few thousand foot
pounds are not uncommon. The contact stress
capacity of space qualified lubricants imposes
serious constraints on usable tooth loads. Hence in
high reduction dries, the designer must consider
dividing the tooth load over a number of gear
meshes. The planetary gearbox is often utilized to
meet these demands.
13.2 Lubrication
The approach to lubricant and material selection for
spacecraft gears must be tempered by consideration of the environment in which they will be
expected to operate.
Conventional oil and grease lubricants are not
suitable for space application since in the presence

space vehicle may not be especially different than

of a high vacuum, normal lubricants will outgas and

the ones found on a shuttle, the duration of the

their usual lubricating properties will be destroyed.


AGMA 911-A94

Some dry compounds, such as graphite, which

exhibit lubricity at sea level become extremely
abrasive in a vacuum. Other lubricants, although
technically usable in a vacuum, will migrate and can
contaminate sensitive optical and electronic equipment. Clean, non-lubricated metal, under the
influence of contact pressure and sliding in a
vacuum, can cold weld. Subsequent separation of
welded surfaces will tear the welded spots apart
leaving severe craters and pits.
Fortunately, certain grease and dry compounds
have been identified, which offer a solution to the
lubrication problem and have been used with
reasonable success. These lubricants fall into two
categories: dry and wet. Each has its own
13.2.1 Application
To be effective, a lubricant must remain in the load
area in order to prevent metal to metal contact on
working surfaces. In ball bearings, the churning
action created by the rolling balls tends to
redistribute lubricant on a continuous basis. In
gears, the tooth action tends to wipe or squeeze the
contact area free of lubricant, especially in slow,
high torque meshes. A planetary gear train will
minimize such lubricant loss since the internal gear
will serve to contain the lubricant, and the circulating
planets will redistribute some lubricant among
working meshes. A successful, long life, wet
lubricated spur gear mesh in space remains to be
13.2.2 Dry lubricants
One dry lubricant which has been in extensive use is
a molybdenum disulphide compound combined
with a carrier. The material is applied to the surface
to be lubricated, and may be subsequently baked,
and finally burnished. The thickness of material
after burnishing is typically in the range of 0.0002
inch (0.005 mm). Since metallic geartooth surfaces
must be separated by a thin residual film of dry
lubricant, the Hertzian contact stress which the
lubricant can carry establishes the load limits for the
gear train. Once the lubricant film is destroyed, gear
failure from scuffing, pitting, and welding can be
expected. As a rule of thumb, gear contact stress
up to 100 000 lb/in2 (690 N/mrr?) has given good
service for low speed, robotic gearing. On meshes

where the pitch line speed may approach 1000

ft/min (5 mls), limiting the contact stress to 35 OOO-


40 000 lb/in2 (240-280 N/mrr?) is prudent. Each

application should be verified by test in vacuum
conditions before being approved for flight.
Some typical solid-film lubricants with space history
are listed in table 17.
13.2.3 Wet lubricants
Both gears and bearings have been successfully
lubricated in space with oil and/or grease for short
periods of time. However, the behavior .of these
lubricants in the space environment imposes some
constraints on the design which must receive
careful attention. Vapor pressure
The vapor pressure of fluid lubricants for space use
is extremely important. Space lubricants typically
have a vapor pressure of I Od to 10-l torr (1.35 x
1o-4to 1.35 x 10 Pa) at 1OOF(38%). Evaporation
and migration of fluid lubricants necessitate the use
of an extensive sealing system which can be
complex and occupy limited weight and space. The
more viscous oils typically have lower evaporation
rates. Surface migration (creep)
Oils vary in their migration characteristics. Mineral
oils like Apiezon C have the lowest migration rates,
while silicone and fluropolymer oils like Versilube
F-50 and Ktytox have the highest. Lubricant creep
may be minimized by the use of flurochemical
barrier films. However, these chemicals must not
enter the load contact area. Greases
Greases are oils that have been thickened by the
addition of either soaps (e.g. stearates) or fillers
(e.g. finely divided silica). In most cases, the
thickening agent is not a good lubricant itself, and
most of the important properties of the grease are
derived from the oil in it. Under vacuum conditions,
the oil will eventually outgas, leaving no lubricant
and sometimes an abrasive thickening agent. A
partial listing of space qualified wet lubricants is
shown in table 18. Rocket lubricants
Rocket gear trains


been successfully

lubricatedwith either oil or engine fuel, as listed in

table 19.

AGMA 91%A94

Table 17 - Candidate solid-film

lubricants for space application


Method of

Lubeco 905


MoS,, PbS & graphite,

+ inorganic binder.

Good load carrying capacity.

Low coef. of friction.
Oil contaminates deteriorate lube.



MoS,, Ag & graphite,

+ ceramic binder.

Ideal for hi temp applications.

Good load capacity. Approx 1000
F cure temp. limits substrate
hardness. Ceramic binder.


Impinged on

MoS,, + ceramic binder.


Moderate load capacity.

Insensitive to oil contamination.



MoS,, & graphite with

sodium silicate binder.

Used on moderately stressed

components. Sensitive to oil



MoS, & graphite, phenolic


Good load carrying capacity.



MoS, ion bombardment

in vacuum.

Good. solid-film coating. Fair to

good load capacity.


- -

Sodium Silicate binder,

MoS, and graphite.

Gear and bearings, deployment


Duroid 5813

- -

Teflon reinforced with

glass MoS,

Self-lubricating material used as

thrust washers, bearing element
separator, etc.




Used as thrust washers, bearing

element separator, etc.


Ion plated


Good load capacity.

x-1 5



* This information is given for the convenience of users of this AGMA information sheet and does not constitute an
endorsement by AGMA of the products named. Equivalent products may be used if they can be shown to lead to
the same results.

13.3 Gear materials for space application

In selecting a gear material, the intended duty,
lubricating media, service life, and environment are
all factors to be considered.
13.3.1 Rocket gearing
The high speed, high power gearing found in
turbopump drives requires more premium grade
material than is used in helicopter and turboprop
units and the suggestions of clause 4 apply.
13.3.2 Spacecraft gearing
Some of the dry film lubricants applied to spacecraft

gearing are hygroscopic and may therefore

promote rusting of the substrate material. To avoid
rust contamination, corrosion resistant steels have
found wide application in steel spacecraft gearing
and structures. For high loads the material can be
gas plasma or ion nitrided. On some designs
precipitation hardening or maraging steels heat
treated to the 44 - 48 HRC hardness range are
used. In general, tooth wear rate is found to be
proportional to the square of the Hertzian stress.
One such material which has found extensive use
for space gears is AMS 5617 in the precipitation
hardened condition.


AGMA 911-A94

Table 18 - Candidate fluid lubricants for space application


Viscous VI,
:Hydrocarbon base oil)


1OOF 210F
138%) (100C)


Space history
and remarks


Vapor pressure 6 x 1O-

torr (8 x 10 -Pa) at

77F (25 C).

Used on OS0 IIINTELSAT III, IV for bearings,

slip ring assemblies,
DC motor brushes, etc.

Versilube F-50
ISilicone oil



Pour point -100F (-75

C). Useful operating
temp. in air is -1 OOOFto
+450F (-75 to +230C).

Used on Gemini, Manner

3 & 4, OGO, Nimbus,
Ranger, Mercury & others.
Used on ball bearings & gears.

Aeroshell7 Soap
thickened diester



Pour point -100F (-75.

C). Useful operating
temp. -100 to +3OOF

Used on Ranger 6 - 9.


to +15OC).

Apiezon c 100
Molecularly distilled
hydrocarbon oil.



Pour point +I 5F (-9C)

Vapor pressure 1O*
ton (1.3 x 1OGPa) at
68OF (20 C).


Braycote 601



Dropping point 1500F

(820 C). Vapor pressure 1 0-gtorr (1.3 x 10v7
Pa) at 38.8F (3.8 C).

Perfluorinated polyether
with tetrafluoroethylene

This information is given for the convenience of users of this AGMA information sheet and does not constitute an
endorsement by AGMA of the products named. Equivalent products may be used if they can be shown to lead to
the same results.

Table 19 - Working fluid lubricants

For lightly loaded applications, plastic gears have
been applied to space gearing. These are used in
combination with stainless steel gears. Plastic
gears giving the lowest wear rate have been MoS,
filled polyimide.
Hard anodized aluminum gears have been used on
a few satelliies, however the coating is subject to
cracking under high loads. Such gears have severe
limitations when run against the same material.
Improved performance will result if the hard

anodized aluminum gear is run against a plastic

gear in the presence of MoS,.



1 ~Space history
~ Titan
Mark 3

Fuel additive
~ Kerosene & 3% Zinc diacoele ~ Delta engine
MIPHI Phosphate
i Gaseous hydrogen
Liquid oxygen

Pratt & Whitney
I Aerojet

AGMA 911-A94

Annex A
Spur gear geometry factor including internal meshes
[The foreword, footnotes, and annexes are provided for informational purposes only, and should not be
construed as a part of AGMA 911-A94, Design Guidelines forAerospace Gearing.]

A.1 Introduction

is dedendum of gear, in (mm);

Many texts dealing with the capacity of spur gears

neglect a discussion of internal gear meshes.
Internal gears are especially important in aerospace applications since planetary reduction
stages are widely used. The definitions of the
involute portion of the internal gear are the same as
the imaginary external gear comprised of the tooth
spaces of the internal gear. The equations
presented herein are written in a form to be
compatible with a computer program.


is addendum of mate, in (mm);


is circular tooth thickness of gear,

in (mm);

To determine the geometry factor, J, for bending

strength of an internal mesh, two general cases
must be considered as follows:
- External gear with internal mate:
- Internal gear with external mate.
The equations presented below are written in a
general form and include signed integers, & and
AM, for the gear and mate respectively. The signed
integers used in the J factor equation permit their
use for both internal and external gears. & and AM
are defined as follows:
= +I for external gear;
= -1 for internal gear;
AM = +1 for external mate;
AM = -1 for internal mate.
The gearis defined, as the member for which the J
factor is desired, whereas the mate is defined as
the mating gear. All calculations are based on a
diametral pitch of 1 .OO and a standard center
distance. The equations apply to a true radius
ground fillet (not generated).

NOTE - tp may be any value so long as:


+ tp + *pnme =n;
is fillet radius of gear, in (mm);

is circular tooth thickness of mate, in

is total backlash, in (mm).
A.2 Geometry factor calculation

NG /2 = pitch radius of gear,

in (mm);


R-(k) b = root radius of gear,

in (mm);


R cos Cp= base circle radius of

gear, in (mm);


N naate/2= pitch radius of mate,

in (mm);


R mate + (AM) amate = outside

radius of mate, in (mm);


Rmate cos q5= base circle radius

of mate, in (mm);


7ccos q5= base pitch, teeth/in


(AM)R + (4) Rmate = center

distance, in (mm);





The following basic data is required prior to starting

the calculation.

+ h%f~

is pressure angle:

is number of teeth in gear;

Nmate is number of teeth in mate;

. ..(A.l)


NO?E - all angles are in radians.


tan @h

is tangent of pressure angle at

highest point of single tooth
contact on involute.


AGMA 911-A94


-(AG) - *i

% =fan&


R;+ [d[R,



. ..(A.7)


for internal or external gears with tangency point

above the base circle, and

tt, is tooth thickness on base circle (may be

cos lY,=$ I A, for all internal gears and for

external gears where the tangency
point is above the base circle diameter

Rx =*

negative for internal gears whereAo =-l), in
inches (millimeters);
rjn is angle which the normal force makes with a
line perpendicular to the tooth centerline at
highest point of single tooth contact, in

r, is the pressureangle at the intersection of the
fillet and involute;
A, is the radius to tangency point of fillet and
gear tooth profile, in inches (mm).

Rx is radius on tooth centerline to point of application of worst load, in inches (millimeters).

To determine coordinates for the center of the fillet


An undesirable condition is a design where the fillet

radius for an external gear is tangent to a radial line
below the base circle. This can happen for small
fillet radii, small pressure angles, and small numbers of teeth. Since the tangency point of the fillet is
below the base circle, there can be no involute from
the end of the fillet to the base circle. The equations
are valid for this condition if a radial lineconnector
is assumed between the base circle and end of fillet.
For internal gears, it is obvious that this condition
can not occur since the fillet must always be at a
larger radius than the base circle. For an external
gear, & = +l and the check for this condition is as
R; -R2
(for external gears only)
2R rr
If the above expression is true for an external gear,
then the tangency point of the fillet is belowthe base
circle diameter and a radial line connector is assumed. To calculateA, for this case, equation A.6 is
used for external gears only. For all internal gears,
or for external gears whose fillet is above the base
circle, the above check is false (for external gears)
and equation A.7 is used to calculate A,.
A, = y/m


for tangency point below base circle diameter, and

cos r, = 1.0000 where the tangency point is

below the base circle diameter


aI= invI;+


1 b




. ..(A.8)

A =a+I,
aa= x -64~)

7 cos A

. ..(A.9)


rf sinA

. ..(A.8)

aa isxcoordinate of fillet radius center, in inches
bb is y coordinate of fillet radius center, in inches
Figure A.1 illustrates some of the nomenclature
used for external spur gears while figures A.2 and
A.3 are for internal gears. Figure A.4 demonstrates
the special case where the fillet center is below the
base circle diameter and the fillet and involute are
connected by a radial line portion.
Afterthe coordinatesof the center of the fillet radius
have been calculated, the next step is to determine
h and t, the height and thickness respectivety of the
tooth at the critical section. Figure A.5 illustrates h
and tfor external and internal gears.

AGMA 911-A94

& space





. -

Figure A.1 - External spur gear nomenclature


AGMA 911-A94


Figure A2 -Angles



locating fillet radius center for internal gear

AGMA 911-A94

k space







I Base

circle I




1 1IIL
ri cp.


Figure A.3 - Internal gear nomenclature


AGMA 911-A94

a ---+jiji


Internal gear

Figure A.4 - External spur gear with fillet

tangency point below base circle
The equations for the calculation of h and t are
different in the zone of the involute and in the zone of
the fillet and are given below for the generalized
radius, A.
In the zone of the involute:
A is radius to the point on involute, in inches

External gear

A lies between A1 and Rx for zone of involute.

OR [(AG < 0) AND (A <Al)]

cosr= -Rb


A is radius to point on fillet, in inches (millimeters).


a =inv r+

1 + (AG)

1 b

In the zone of the fillet:


Figure A.5 - Definition of built-in

section of tooth

-(AG) fb


[(AG< 0) AND (A >A1 11

A lies between Al and R, for zone of fillet.

AGMA 911-A94

[ ~;!+?3;$]
NOTE-if (A1sAsRb )thena=a (radiallineportion)


+;AG> NG- wo;]




l+@G z;n
2 9

(&) a]


Similar equations may be developed forthe case of

a generated fillet.
After defining the general equations to calculate h
and t at any point, it remains to determine the
location where the stress will be the highest. The
classical method which dates from the Lewis
equation was to assume that the tooth had an
inscribed parabola with the apex at the load point
and which was target to the tooth as shown in
figure A.6.

ences other than pure bending which are functions

of h and t:
- The axial component of the normal tooth load
which produces a compressive stress that
subtracts from the bending;
- The stress concentration factor which is a
function of h and t.
Therefore, the point at which hand tact is defined as
the point which produces the highest total stress
and not the highest bending stress. Since all the
influencing factors are contained in the geometry
factor, J, the object is to determine h and t where J is
J is defined by:


cos 4n
Kt---- cost [ x

tan 4%

. ..(A.l6)


Kt =H + [+]

. ..(A.l8)


where H, M, and L are given in table A.1 .

Table A.1 - H, L, and M for use in Kt
(from linear extrapolation of
Dolan and Broghamer method)

Figure A.6 - Constant stress parabola

angle $





* Original Dolan and Broghamer values

Since a parabola has the property of constant

bending stress, section B-B, which is at a point of
tangency between the parabola and tooth profile will
have the maximum bending stress. Note at any
other section such as A-A, the stress would be the
same as at section B-B only if the tooth were of
thicknesst~. Howeverintheactualtoothatsection
A-A has more metal outside of tM and therefore
has lower stresses. This classical assumption is

Assuming that H, M, and L are straight line functions

of the pressure angle, they may be expressed as

based on constant bending stress. However,in the

procedure is used utilizing the equations for h and t

modern definition of tooth stress, there are influ-

which have been previously defined as a function of

H = 0.3255- 0.4167Cp
L = 0.3318 - 0.52094
M = 0.2682+ 0.5209$I
where $ is in radians.
To find the point where J is a minimum, an iterative


the general radius, A. A simple method is to divide

the section between R, and R, into fBed number
of intervals, n , and to start at some point such as
R, . J is then successively calculated using a new
value of A each time defined by:
. ..(A.19)

Figures A.7 through A.9 are for external gears with

internal mates for gears of varying pressure angles,
while figures A.10 through A.12 are for internal
gears with external mates. These curves may be
used for preliminary design purposes prior to the
detailed Jfactor calculations using the actual tooth

J will continue to diminish until a minimum is found,

at which point the next value of J will increase. The
accuracy of the answer will depend on the magnitude of the interval A. With the advent of high speed
computers, the calculation of Jby the above method
is easily accomplished. Figures A.7 thru A.12 are
plots of the geometry factors for internal/external
gear meshes for gears with standard proportions.

Figures A.7 through A.12 have been developed

using the equations in thisannex, with the input data
shown in tables A2 through A.7. Note that all values
shown are for diametral pitch = 1 .OO. Maximum
value a = 1 .O was used for the addendum and
Br,=O.O4 for the the backlash; minimum value
d = 1.24 was used for the dedendum and t = 1.5508
for the tooth thickness.



AGMA 911-A94





Number of teeth in external gear



Figure A.7 - Geometry factor for standard proportions, external gears with internal mates,
addendum = 1 .OO,dedendum = 1.24, tooth thickness = 1.5508, backlash = 0.04, (I = 20
Table A.2 - Fillet radius used to generate J factor curves in figure A.7
Fillet radius, inches
Number of external teeth




of internal teeth



AGMA Qil-A94



Number of teeth in




Number of teeth in external gear



Figure A.8 - Geometry factor for standard proportions, external gears with internal mates,
addendum = 1 .OO,dedendum = 1.24, tooth thickness = 1S508, backlash = 0.04, $ = 22.5
Table A.3 - Fillet radius used to generate J factor curves in figure A.8
Number of internal teeth



Fillet radius, inches

Number of internal teeth














AGMA 911-A94



% t$ -52

Number of teeth in external gear



Figure A.9 - Geometry factor for standard proportions, external gears with internal mates,
addendum = 1 JO, dedendum = 1.24, tooth thickness = 1.5508, backlash = 0.04, + = 25O
Table A.4 - Fillet radius used to generate J factor curves in figure A.9
Fillet radius, inches
Number of external teeth


Number of internal teeth





















AGMA 911-A94




8L 5 .56


.48 '


80 100
Number of teeth in internal gear



Figure A.10 - Geometry factor for standard proportions, internal gears with external mates,
addendum = 1 .OO,dedendum q 1.24, tooth thickness = 1.5508, backlash = 0.04,$ = 20
Table A.5 - Fillet radius used to generate J factor curves in figure A.10
Fillet radius, inches
Number of internal teeth



Number of external teeth





























AGMA 91%A94




Number of teeth in
matinn external near




17 20
\ \




.52 ~~



80 100
Number of teeth in internal gear



Figure A.11 - Geometry factor for standard proportions, internal gears with external mates,
addendum = 1 .OO,dedendum = 1.24, tooth thickness = 1 S506, backlash = 0.04, $ = 22.5
Table A.6 - Fillet radius used to generate J factor curves in figure A.11

Fillet radius. inches

Number of internal teeth

of external teeth










AGMA 911-A94

3 5 .62 aao

= g.60









Number of teeth in internal gear

Figure A.12 - Geometry factor for standard proportions, internal gears with external mates,
addendum = 1.99, dedendum = 1.24, tooth thickness = 1.5599, backlash = 0.94, $ = 25
Table A.7 - Fillet radius used to generate J factor curves in figure A.12

Number of internal teeth


Fillet radius, inches

Number of external teeth


-- I


AGMA 911-A94

Annex B
Gearbox Test and Mission Requirements
[The foreword, footnotes, and annexes are provided for informational purposes only, and should not be
construed as a part of AGMA 91 l-A94, Design Guidelines for Aerospace Gearing.]
For gearbox design and test purposes values of
B.l Introduction
power and speed are needed for each phase of
Mission profiles are established to determine load,
flight. Figures B.2 and B.3 show typical power and
speed, operating time, and environmental requirespeed values versus time for both a fixed wing and a
ments for gearbox operation during service. The
rotary wing application.
designer uses these parameters to size the gearbox
while testing is performed to assess the integrity of
---the design and to uncover any unmanifested
problems. Gearbox testing is performed in test rigs
where possible, and finally in the actual application
for certification.

B.2 Mission Requirements

Missionsvary depending on the customer, military or
commercial, and on the type of aircraft/ engine such
as propeller driven, helicopter and turbojet engines.
The basic phases of flight are shown, for a
commercial aircraft, in figure B.l which displays
altitude versus time. The definition for each
numbered phase is listed as follows:




. .










% lime

Figure B.2 - Commercial application,

turbofan aircraft

1 - Takeoff 4 - Altitude change 7 - Landing

8 - Ground
2 - Climb 5 - Descent
3 - Cruise 6 - Approach
Other maneuvers include Holding Patterns while
above airports and Go-Arounds during landing
attempts. For some military aircraft Loitering is a
phase that may consume 25% or more of the
mission time.







% Time

Figure B.3 - Military application, helicopter


Figure B.l - Commercial application,

phases of flight

For actuator gearboxes, the highest loads are

usually experienced at the lowest speeds and in
many cases these loads occur at zero speeds (zero
powertransmitted). These loads are referred to as
holding loads, stall ioads, or limit loads. They
are def&f by the customer for a specific applica-


AGMA 911-A94

tion, and hence, they become a major criteria

governing the design.
Commercial applications are often more stringent
than military in terms of gearbox loading and hours
of use. Military aircraft, however, operate in a more
severe environment which may shorten gearbox fife
for reasons other than load.
Figure B.4 represents a 150 hour certification test for
a helicopter engine that containsa Power Reduction
Gearbox that must be certified also. The test is
designed to be more severe in terms of load than an
equivalent number of hours of field operation. For
example, if power levels below 90 % do not have a
significant effect on the accumulated fatigue damage of the gearing, then only 2-l/2 % of the mission
time shown in figure B.3 would be applicable to
reducing fatigue life, a matter of l-1/2 minutes. By
comparison, 50 % of the time, 4 500 minutes, is
spent at or above 90 % power in the 150 hour
qualification test. This illustrates that sufficient
operational hours can be accumulated during a
certification test to reveal long term field related
operational problems.
Figure 8.5 shows power usage spectrumsfortypical
small and large military helicopters. In general, as
the size of the helicopter increases, the percent time
near the top power levels increase.

B.3 Testing
Testing is performed to assess the integrity of the
gearbox design and is a requirement for aircraft and

engine flight certification. The endurance and

environmental testing outlined in the following
sections is extracted from the Army Aviation Specification AV-E-9593D, 1934 for turboprop and
turboshaft engines. Numerical values are listed to
show the stringency of the test requirements which,
when completed successfully, will lead to safer
operation of the aircraft. Limit and ultimate load
testing are included to cover actuator testing.








Figure 8.5 - Milita~~~l,icopters, large and

8.3.1 Types of Testing
8.3.1 .l Development Testing
Development testing is required to assess the
capability of a transmission design prior to qualification or flight. Certain tests, as outlined below, are

Repeat the 5-hour cycle 25 times

Figure B.4 - 150 Hour engine certification test



% Time

AGMA 911-A94

required to establish and develop basic operational

capabilities such as lubrication system tests and
gear pattern development tests. Other development
tests are conducted to prove out structural integrity,
search out weak links and gain confidence in the
design. The ultimate goal is to verify that the design
requirements for reliability and maintainability have
been achieved by the time the transmissions are put
into service.
- Lubrication Systems Tests - Testing at no load
and varying speeds up to overspeed conditions.
Bearing temperatures are monitored for
over-temperature conditions. Oil quantity, pump
flow rate and pressure are established. Then
power is applied for development of cooling
system requirements;
- Gear Pattern Development Tests - Gears are
tested to determine load distribution at various
power conditions. Visual techniques as well as
strain gaging are used to determine load diitribution as a function of applied load. Machine
grinding corrections are used to assure full
contact at the desired power level and to correct
for deflections. After testing, final machine
corrections are added to the production manufacturing procedure;
- Over-torque Development Tests - Tests are
conducted at over-torque conditions using accelerated mission profiles to substantiate a higher
equivalent operating life and to demonstrate
non-catastrophic and fail-safe modes through
the use of different inspection and detection
techniques. Over-torque testing often accelerates failure modes. This is beneficial in a
development test program so that design or
manufacturingchanges can be made in a timely
- Bench Tests - Used to substantiate systems
which have loads induced from external sources,
i.e., a shaft and housing which support a propeller
or a rotor;
- Vibration Surveys - Vibration surveys are
conducted to determine resonances in the operating range. Mode shapes for the resonant
frequencies are found by holographic interferometry, rap testing with accelerometers, or
other methods.

B.3.1.2 Endurance Testing

Endurance tests are performed to assess the
integrity of the gearbox design and are usually
performed with maximum parameter conditions
acting simultaneously.
- Accessory Gearbox Drive Test - 300 hour
endurance/ 600 starts. Simultaneous operation
of all drives with each drive subjected to the
maximum permissible torque or power rating
including the maximum overhung moment and
misalignment angle of each accessory. A vibratory and resonant search from engine idle up to
125 % of maximum speed under varying loads.
Tests to be conducted at maximum oil inlet
temperature and minimum oil flow. The same
gearbox is to be used for the Static Test;
- Accessory Gearbox Static Test 150%
maximum static torque on all accessory pads
simultaneously and with the starter loaded to 250
% of max. starting torque for five seconds;
- Engine Output Shaft Drive Spline - Demonstrate ability to absorb thrust In either direction
equal to 20% of the circumferential force acting
on the output drive shaft spline at the maximum
continuous torque rating. Perform endurance
testing at maximum allowable misalignment and
at a speed of 1.15 times the maximum continuous
rating. Demonstrate a continuous operating
capability of 1.2 times the maximum continuous
torque rating;
- Oil Interruption Test - Operate engine for 30
minutes at intermediate power (a rating somewhat higher than maximum continuous power)
with air only at the oil pump inlet. Then with oil
supplied, operate the engine for 30 more minutes.
There shall be no damage to any components as
a result of the test;
- Windmilling Test (backdriving) - 8 hours
continuous operation without component damage or excessive loss of lubricant:
- Low Cycle Fatigue - A test cycle from engine
idle to maximum power and back to idle lasting
approximately 5 minutes with a shutdown every
15 minutes. Repeat cycle for a total of 3 750
cycles. The minimum LCF life of the components
shall be 15 000 cycles.


AGMA 911-A94

B.3.1.3 Environmental Testing

A number of environmental requirements exist for
the safe operation of airborne gearboxes. Some of
the Military requirements and reason for testing, as
outlined in MlL-STD-31 OC, are listed as follows:
NOTE - Numericalvalues are not listed here as they
are extensivelycoveredin the MIL-STD.

- Low Pressure (altitude) - Gearbox venting, oil

leakage and effect on heat convection and
conduction characteristics;
- High Temperature - High temperature storage
or service conditions;
- Low Temperature - Loss of resiliency of gasket
materials, congealing of lubricants and tight
meshing of gears;
- Solar Radiation - Damaging effects on natural
rubber and plastics;
- Attitude -. Insure proper oil scavenging for all
flight maneuvers and for leakage during transportation;
- Rain, Humidity and Fungus - Component
damage resulting from wet and warm atmosphere;
- Salt and Sand - Component corrosionlerosion;
- Flame - Fires during storage or operation;
-Vibration and Noise- Performance degradation
or component breakage and airport operating
- Shockor Impact-Operation and transportation
- Combined Effects - Where multiple conditions
may exist simultaneously i.e. Altiiude - combining low pressure with high temperature. Space combining low pressure and temperature with

induced dynamics. Testing duration is usually from

50 to 200 hours. Drive shafting, cooling systems, tail
and intermediate gearboxes, and accessory drives
are also tested.
After ground testing has shown that no major
development issues exist, the transmission is tested
in flight. Flight and ground testing continue simultaneously until qualification is complete. The objective
of the flight test program is generally fourfold:
1) Structural substantiation;
2) Demonstration of handling qualities and the
automatic flight control system;
3) Verification of the propulsion/drive system;
4) Confirmation of the mission equipment.
Structural substantiation testing establishes and
expands flight envelopes including evaluation of all
attitudes and conditions of flight. This includes
extreme maneuvers and maximum and minimum
loads on all major components. Handling qualities
confirms basic data for the controls and control
system characteristics and evaluates control
reliability and fidelity. Drive system testing confirms
structural integrity and measures flight loads
required to establish component life. Mission
equipment testing includes instrumentation, navigation and communications systems quality, human
factors criteria, and related systems capabilities and
B.3.1.5 Actuator qualification testing Flight Testing

Testing requirements are determined largely by the

part reliability requirements. Test programs often
include, but are not limited to the following:
- Baseline performance testing;
- Thermal vacuum testing;
- Vibration/shock testing;
- Humidity testing;
- Lie testing;
- Structural load testing.

Prior to conducting flight testing, development

testing has usually been completed or is well along
in schedule. Some qualification agencies require tie
down testing of a helicopter with the complete
dynamic system installed prior to flight test. The
helicopter is tied to the ground and the rotors
operated to full capabilii levels to simulate flight.
The transmission receives full torque as well as rotor

Baseline performance runs normally are conducted

before and after each test. Structural load tests are
performed to validate the stress model and to verify
the load capacity. Thermal vacuum tests are often
expensive, but they are a reliable method of
determining the mechanism characteristics under
environmental extremes. Life testsoften exceed the
operational requirements by a factor from 1.5 to 4.0.

NOTE - Referto clause 6 for a more detaileddiscussion.


AGMA 911-A94

This is done to ensure that all gearboxes meet the

Large systems such as a robotic manipulator often
can only be qualified by testing individual components due to the large size of the assembled unit.
Guidance for testing can be found in Environmental
Test Methods and Engineering Guidelines, MILS-1
B.3.2 Test cell and installation


The condition of test cell apparatus is paramount to

the successful completion of a program.
acquisition should be complete and should be
carried out in a timely manner.
The following
checklist may serve as a guideline for complete and
orderly accrual of data:
a) Pm-test:
- Calibration of test apparatus and instrumentation;
- Means of recording test cell temperature, humidi and atmospheric pressure;
- Inspection records of all gearbox components;
- Test log for proper certification of test data;
- Test parameters and sequence to be cornmensurate with expected mission requirements;
- List of all test equipment and facilities;
- Adherence to all Safety requirements:
- Review normal and emergency shutdown
- Clean room conditions when required.
2) Startup:
- Clean chip detectors
- Heat or cool test item and oil;
- Operate scavenge pumps;
- Operate oil pump;
- Apply load with load cell;
- Check temperature, pressure, speed, vibration, flow indicators and any special test equipment.
3) During Testing:
- Record all data for the established test sequence;
- Proper documentation of all problems and
4) Post-test:
-Verification of test apparatus and instrumentation calibration;

- Visual, dimensional, and non-destructive inspection record of all gearbox components;

- Spectrometric analysis of the oil and chip detectors;
- Evaluation of all test data in preparation for a
final report.

8.4 Gearbox

Test Rigs

Accessory and actuator gearboxes may be loaded

by the actual accessories or by using waterbrakes or
other load absorbing devices on the mounting pads.
Electric motors coupled to step up gearboxes are a
common method for providing power. Variable
attitude rigs are used to evaluate the effects of oil
churning and scavenging during a simulated mission. In this rig the whole gearbox is mounted in a
stand capable of rotation around two axes, pitch and
roll, during operation.
Main reduction gear driie systems would require
such large loading as well as power absorption
devices as to make this method of testing quite
expensive. The four square system eliminates the
need for a load dissipating device and reduces the
size of the loading device considerably. The system
is closed loop in that input and output shafts are
coupled together.
Figure B.6 illustrates a four square system that uses
two gearboxes arranged back-to-back.
One end of
the two piece output (low speed) shaft is held fiied
while the other end is rotated to take up backlash
and then rotated until the desired torque is reached.
The two halves are then locked together and
connected to the prime mover through a clutch. Slip
rings are used to measure torque, the clutch assures
smooth acceleration of the rig tospeed, and the
power required by the prime mover need only be
enough to overcome the system friction which
amounts to approximately 2% of the gearbox design
point power. On one side of the rig, the gears will be
loaded as in normal operation (test). On the other
side (slave), the gears will be loaded on the coast
side. Thrust load direction, for helical gears, and oil
jet location may require special consideration, for
the slave gearbox, when designing this type of
system. The advantage of this system is the minimal
amount of new hardware required, i.e. input and
output shafts. The disadvantage is that the load is
not variable during testing.

AGMA 911-A94


Test side

Slave side

Prime mover


Output (low speed) shaft

Figure B.6 - Test rig arrangement, back-to-back
The top view of another type of four square rig is
shown as in figure 6.7. In this arrangement a single
planetary reduction gearbox is coupled to a low
speed gearbox on the left side and a high speed
gearbox on the right. Slip ring assemblies measure
both input and output torque. The gears numbered
Na and Nb have the same values in both the
reduction gearbox and the rig. The low speed shaft
can be reduced further in speed by choosing 3Na or
4Na etc. The Input Torque Drive is located on the
low speed shaft and incorporates splines that are
hydraulically loaded to provide the desired torque.
The advantages of this configuration are a single
test gearbox is required with all gear teeth loaded
and rotating as in normal operation. Also, loading
may be varied during running. The disadvantage is
the extra gears and attendant bearings that must be
designed for each side of the rig and the increase in
size of the prime mover to overcome system friction.

B.5 Production dynamic testing


test for larger complex gearboxes/transmissions.

The complex testing may be performed in test rigs as
described in B.4, or on the next higher assembly
dynamic test as in the case of turbine engine
accessory or speed reduction gearboxes.
B.5.1 Rig acceptance tests
These tests should be described in detail by
acceptance test plans (ATPs). The ATP should, as a
minimum, cover:
a) Pretest procedures;
b) Startup procedures;
c) Load spectrums;
d) Data to be recorded and documented;
e) Inspections to be performed during test;

9 Shut down procedures;

g) Post test inspections to be performed.
8.5.2 Next assembty subsystem testing

After each air vehicle gearbox/transmission undergoes final assembly, it is prudent to substantiate
design performance by a dynamic test. The test

Written procedures should be used to define the test

parameters as listed in B.5.1 for the gearbox/transmission being tested. This type of testing is cost
effective, but does not have the risk of high

performed can be simple or complex, i.e. spin test for

disassemblycosts when infant mortality failures are

small accessory gearboxes or a mini qualification



AGMA 911-A94

Test gearbox

Slip ring

Slip ring





Low speed gearbox


fluid in

Figure 8.7 - Test rig arrangement, variable torque


AGMA 911-A94

This page is intentionally blank.


AGMA 911-A94

Annex C
References and Bibliography
[The foreword, footnotes, and annexes are provided for informational purposes only, and should not be
construed as a part of AGMA 911-A94, Design Guidelines for Aerospace Gearing.]
1. Anderson, N.E., Loewenthal, S.H.: Comparison
of Spur Gear Efficiency Prediction Methods, Advanced Power Transmission Technology, NASA CP
2210, AVRADCOM TR 82-C-l 6, June, 1981.

9. Drago, R. J., Then AnalicalExperimentalEvaluation of Resonant Response in High Speed, Lightweight, High/y Loaded Gearing, ASME Paper
8O-C2/DET-22, August 1980.

2. Anderson, N.E., Loewenthai, S.H., Black, J.D.:

An Analytical Method to Predict Efficiency of Aircraft Gearboxes, ASME Journal of Mechanical Design, v. 108, Sept., 1986.

10. Lisp, T. C. and Zakrapak, J. J., ModalAnalysis of

Gear Housing and Mounts, Seventh International
Modal Analysis Conference, Jan. 30,1989.

3. Anderson, N.E., Loewenthal, S.H.: Efficiency of

Nonstandard and High Contact Ratio Involute Spur
Gears, ASME paper 84-DIET-172, presented at the
Fourth ASME International Power Transmissioin
and Gearing Conference, Cambridge, Mass., Oct.
4. Harris, T.A.: Rolling Bearing Analysis, Wiley, New
York, 1966, pp446-450.
5. Kleckner, R.J., and Pirvics, J.: High-Speed Cy/indrical Roller Beating Ana&sis (CYBEAN) Users
Manual, SKF Report No. AL78P023, SKF Industries, Inc., (NASA Contract No. NAS3-22807), July,
6. Kleckner, R.J., Dyba, G.J., and Ragen, M.A.:
Spherical Roller Bearing Anal&is (SPHERBEAN)
Users Manual, SKF Report No. AT81D007, SKF Industries, Inc., (NASA contract NAS3-22807) Feb.,
7. Hadden, G.B., Kleckner, R.J., Ragen, M.A., and
Sheynin, L.: System lnciuding Ball, Cylindrical, and
Tapered Roller Bearings (SHABERTH) Users Manual, SKF Report No. AT81D040, SKF Industries,
Inc., (NASA Contract No. NAS#-22690) May, 1981.
8. Akin, L.S., Townsend, D.P.: Lubricant Jet Flow
Phenomena in Spurand Helical Gears with Modified
Addendums-for Radially Directed individual Jets,
NASA TM 101460, AVSCOM TR 88-C-034, presented at the Ffih International Power Transmission
and Gearing Conference, Chicago, Illinois, April,

Il. Koster, W-P., Surface /ntegrity of Machined Materials, Technical Report, AFML-TR-74-60, April
12. Fuchs, H.O., Shot Peening Stress Profi/es.
13. Lauchner, E., Westech Presentation, March
1974, Northrop Corporation, Hawthorne, California.
14. Ahmad, Aquil, Eaton Corporation.
15. Daly, J., A Concept for Using Controlled Shot
Peening in Original Gear Design, American Gear
Association Technical Paper
16. Horger, 0-J. and Lipson, C., Automotive Rear
Ax/es and Means of improving Their Fatigue Resistance, American Society for Testing and Materials,
Technical Publication No. 72, 1947.
17. Lowenthal, S.H.; Design of Power Transmission
Shafting, NASA Report, RP-1123.
18. Townsend, D-P., and Zaretsky, E.V., Effect of
Shot Peening on Su/face Fatigue Life of Canburized
and HardenedAlSl9310 Spur Gears, NASA Technical Paper 2047,1982.
19. Dudley, D-W., Handbook of Practical Gear Design, McGraw-Hill, Inc., 1984.
20. Prevey, P.S., X-Ray Diffraction Residual Stress
Techniques, Metals Handbook, Ninth Edition,
Vol.1 0, ASM International, Ohio, 1986.
21. A. Gerve, B. Kehrwald and L. Wiesner, T. W.
Conlon and G. Dearnaley, Materialscience and Engineering 69 (1966), pp. 221-225.
22. G. Hubler, I.L. Singer and C. R. Clayton, Materia/s Science and Engineering 69 (1985), p. 203.