See

discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/223573607

A study of spur gear pitting formation and life

prediction
ARTICLE in WEAR DECEMBER 2004
Impact Factor: 1.86 DOI: 10.1016/j.wear.2004.08.005

CITATIONS

DOWNLOADS

VIEWS

33

271

278

2 AUTHORS:
Kubilay Aslantas

Suleyman Tasgetiren

Afyon Kocatepe University

Afyon Kocatepe University

53 PUBLICATIONS 172 CITATIONS

60 PUBLICATIONS 165 CITATIONS

SEE PROFILE

SEE PROFILE

Available from: Kubilay Aslantas

Retrieved on: 14 August 2015

Wear 257 (2004) 11671175

A study of spur gear pitting formation and life prediction

K. Aslantas , S. Tasgetiren
Technical Education Faculty, Afyon Kocatepe University, Afyon, Turkey
Received 1 March 2004; received in revised form 9 August 2004; accepted 9 August 2004
Available online 28 September 2004

Abstract
In this study, a numerical prediction on pitting formation is carried out in spur gear made from austempered ductile iron. General twodimensional rolling sliding contact situations are considered for the development of the analytical model. The problem is assumed under
linear elastic fracture mechanics and the finite element method is used for numerical solutions. Mixed mode stress intensity factors KI and
KII for cyclic loading are evaluated and related to crack extension by a Paris-type equation. The maximum tangential stress criterion is used
to determine the crack-turn-angle during crack propagation under cyclic loading.
A series of experimental study is also carried out to determine the pitting formation life. Test specimens were first austenitized in salt bath at
900 C for 90 min after which they were quenched in salt bath at 325 and 425 C, for 60 min. A comparison is carried out between numerical
and experimental results.
2004 Elsevier B.V. All rights reserved.
Keywords: Spur gear; Austempered ductile iron; Pitting; Fatigue; Finite element analysis

1. Introduction
Rolling contact fatigue cracks are one of the most important problems for the gear industry. Under rolling contact,
various surface damages (pitting, spalling and cracking) occur and also cracks develop in the machine units, thus leading to loss of serviceability of the machine part. Pitting failures are formed in all rolling pairs such as gear, rail-wheel
etc.
Austempered ductile iron has recently been developed for
an increasing number of the engineering applications such as
gears, crankshaft, connecting rods and other heavy machinery and transportation equipment because they offer excellent
combinations of high strength, ductility, toughness, fatigue
strength and wear resistance over other grades of cast irons
[1]. Mechanical properties of austempered ductile irons depend on the heat treatment conditions, graphite size and the
distribution of defects in the microstructure.
Modelling studies of crack propagation and spall formation based on the fracture mechanics have been carried out

Corresponding author. Tel.: +90 272 228 1235; fax: +90 272 228 1235.
E-mail address: [email protected] (K. Aslantas).

0043-1648/$ see front matter 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.wear.2004.08.005

by researchers in wear studies. Aslantas and Tasgetiren [2]

studied the behaviour of a subsurface-edge crack subjected
to moving normal and tangential loads. The problem is considered under the linear elastic fracture mechanics theory. KI
and KII stress intensity factors at the crack tip are computed
for different load positions and different load application angles. Tasgetiren and Aslantas [3] also studied the growth behaviour of surface crack having three different angles to the
surface for different load positions and friction coefficients.
The Hertzian contact model was considered for loading conditions.
Chue and Chung [4] have analysed the mechanism of pitting caused by rolling contact by using fracture mechanics
approach. The authors considered the initial crack length,
crack angle, contact force, friction, strain hardened layer and
hydraulic pressure of trapped fluid acting on the crack surface
as variables of the analyses. Keer and Bryant [5] attempted
to simulate a crack propagation mechanism for pitting. In the
study, pitting formation life is calculated and the effect of the
contact friction, lubricant pressure and friction between the
crack faces are discussed. The authors determined the fatigue
life by increasing the cracks length and stress intensity factor
using an analytical method.

1168

K. Aslantas, S. Tasgetiren / Wear 257 (2004) 11671175

Sheppard et al. [6] have considered an elastic half-plane

containing a surface-breaking crack normal to the free surface, subjected to loading by uniform tractions over a given
length of its surface. Stress intensity factors are computed
for various crack lengths, friction coefficients and ratios of
applied tractions. Hills and Comninou [7] studied an elastic
half-plane subjected to a uniform pressure over part of its
surface. A normal edge crack is assumed to lie at the edge
of the loaded region. The effect of Coulomb friction between
the crack faces is considered, and the various possibilities involving stick and slip regions along the crack are examined.
Stress intensity factors are obtained.
Magalhaes et al. [8] studied surface and subsurface crack
initiation in austempered ductile iron discs. The authors used
several auxiliary surface analysis techniques. Magalhaes and
Seabra [9] analysed wear and scuffing properties of gears
made from austempered ductile iron. In the study, FZG tests
were performed for five gear sets and three different austempering temperatures were used.
In this study, a numerical model for the prediction of pitting formation life is carried out in spur gear made from
austempered ductile iron. For this purpose, a subsurface crack
is considered first. Then the crack growth is modelled by using the linear elastic fracture mechanics principles. The analytical model developed is considered as in the general twodimensional rolling sliding contact situations. Mixed mode
stress intensity factors KI and KII for cyclic loading are evaluated and related to crack extension by a Paris-type equation.
The maximum tangential stress criterion is used to determine
the crack-turn-angle during crack propagation under cyclic
loading. Several experimental studies are also carried out to
characterize the material and to determine the material properties to be used in the numerical model.

E =

2Ep Eg
Eg (1 p2 ) + Ep (1 g2 )

R =

R p Rg
Rp + R g

(2)

The pinion used in the study is made of surface hardened (58

HRc) AISI 8620 steel whose elasticity modulus is 207 GPa.
The gear material is the investigated austempered ductile iron.
The elasticity modulus and Poissons ratio of the studied ADI
are obtained experimentally as 170 GPa and 0.25, and does
not change with applied heat treatments. The geometrical
parameters of the gear and the pinion are given in Table 3.
The radius of the equivalent cylinders are obtained as 8.75
and 21 mm for pinion and gear, respectively. As a result, the
equivalent elasticity modulus E and the equivalent radius R
are obtained as 201.8 GPa and 6.17 mm.
The normal force is a function of the maximum Hertz
contact pressure P0 :
2P
c

(3)

Distribution of the normal contact pressure in the contact

area can be written as:

2.1. Contact analysis

When two solids in relative motion are brought into contact, the resulting normal and tangential forces at the contact
region are transmitted from a surface to its pair. The gears
are one of the examples of solid bodies used for the transmission of forces. For the purpose of easier consideration of the
gear contact parameters in numerical computations, it is very
useful to replace the complicated gear pair geometry with an
equivalent model of two cylinders [10]. Equivalent cylinder
model for gear pair is shown in Fig. 1.
According to the Hertz elastic theory [11], when two cylindrical elements move on each other, the half-length of the
contact area can be determined from:
4PR
E

determined using Eq. (2) in which Ep and Eg are the elasticity

modulus of the pinion and the gear, respectively and Rp and
Rg are radius of the pinion and the gear.

P0 =

2. Numerical procedures

c2 =

Fig. 1. Equivalent radius for gear and pinion.

(1)

where P is the applied normal force, E the equivalent elasticity modulus and R is the equivalent radius. E and R are

1/2

P0 (c2 x2 )
p(x) =
c

(4)

Table 1
Chemical composition (wt.%) of the gear material
C
Si
Mn
P
S
Mg
Cr
Ni
Mo
Cu
Al
Ti
Sn

3.40
2.69
0.19
0.02
0.01
0.044
0.04
0.73
0.23
0.87
0.015
0.004
0.007

K. Aslantas, S. Tasgetiren / Wear 257 (2004) 11671175

1169

Fig. 2. (a) The location of the nodes used to calculate the stress intensity factors; (b) six-node quarter point isoparametric elements at the crack tip.

2.2. Calculation of the stress intensity factors

A mixed mode (mode I and II) stress intensity factor solution is developed for the general crack geometry using the
finite element methodology. In this study, the displacement
correlation method is used for calculation of the stress intensity factors. The method is one of the most popular methods
to calculate stress intensity factors by numerical techniques.
After the finite element or boundary element solutions for
cracked structure are obtained, nodal displacement values of
nodes 2, 3, 4 and 5 (Fig. 2) are determined. These displacements are utilized for the calculation of the stress intensity
factors.
The crack face displacements in both opening and sliding
modes are related to the stress intensity factors for mode I
and mode II. The opening mode KI and the shear mode KII
are calculated by [12]:
KI =

G
+1

2
[4(v2 v4 ) + (v5 v3 )]
L

(5)

G
2
KII =
[4(u2 u4 ) + (u5 u3 )]
+1 L

 3 4, for plane strain

=  3
 1+ ,
for plane stress

(6)
(7)

where L is the element length, the Poissons ratio, and ui

and vi are the nodal displacements in the x and y directions,
respectively. In the present study the angle of crack growth
associated with each load position and crack length is found
from the maximum tangential stress theory using the current
load steps SIF as:



2
1
K
K
I
I
= 2 tan1

+ 8
(8)
4 KII
KII
2.3. Finite element model of the gear
The finite element method has been widely used to solve
problems in linear elastic fracture mechanics. The main diffi-

Fig. 3. (a) Finite element mesh; (b) the subsurface crack in gear tooth and surface loads effect on tooth surface.

1170

K. Aslantas, S. Tasgetiren / Wear 257 (2004) 11671175

Fig. 4. Microstructure of the ductile iron material: (a) as-cast; (b) austempered at 325 C for 60 min.

culty in these calculations stems from the occurrence of infinite stresses at the crack tip. An eight-node isoparametric element is applied in a plane strain configuration that represents
the contact between the gear and pinion. In the mesh around
the subsurface crack, shown in Fig. 3a, six-node isoparametric elements with mid-side nodes adjacent to the crack tip are
used.
The gear is modelled as a half cylinder. The displacements at the bottom are fixed in the x and z directions. The
effect of the pinion is shown as elliptic loading according
to the Hertz theory. The position of the load is shown with
d which is the distance between the centre of the crack
and the maximum load in the elliptic variation. In Fig. 3b,
the depth of the subsurface crack is shown with h. The
depth is determined by stress analysis for uncracked finite
element model. Von-Mises yield criteria is used for the stress
analysis.
The simulations were performed with the FRANC2D [13]
finite element code. Among the variety of capabilities, a
unique feature of FRANC2D is the ability to model a crack
in the structure; 1080 elements and 3311 nodes were used for
the entire mesh.

3. Experimental study
3.1. Material
The material used in the present study is a nodular graphite
cast iron alloyed with Ni and Mo. The chemical composition of the material is given in Table 1. The microstructure of the as-cast material is shown in Fig. 4. The as-cast
material has 80% pearlitic and 20% ferritic structure. Gear

Table 2
Heat treatment and mechanical properties of the gear material
Sample Austempering
temperature ( C)

Hardness
HRC

y (MPa) UTS (MPa) (%)

A
325
B
425
As-cast

42
37
23

804
565
493

1228
862
780

4.3
9.2
7.1

samples for pitting testing were machined from Y blocks

of thickness 50 mm. In addition, metallographic samples
were prepared according to the standard procedures. The
gear material has 95% nodularity and the nodule count is
200 nod/mm2 .
Three different samples are studied in the experiments.
One as-cast, one (A) austenitized at 900 C for 90 min and
quenched in salt bath at 325 C for 60 min. Austenitization
of the last specimen (B) is the same as the second specimen
but this is quenched at 425 C for the same duration. The
mechanical properties of the specimens are given in Table 2.
The values of the Table 2 are the mean of three samples
results.
3.2. Pitting tests of gears
Pitting tests of the gear pairs are carried out on a FZG
(Forschungsstelle fur Zahnrader und Getriebebau) test machine which has different gear centre distances. The FZG test
machine is back-to-back spur gear equipment with a closed
power circuit [14,15]. Test pinion and test gear are mounted
on two parallel shafts which connect the drive gear stage with
the same gear ratio (Fig. 5). While not in motion, the transmission can be blocked by a locking pin. By twisting of the
load clutch using weights on the load lever, the desired test
torque is applied, which is indicated at the torque measuring
clutch.
Three different tests are performed using lubricant (Mobil
Gear SP 150 with kinematics viscosity vk = 150 mm2 /s). The
geometrical parameters used in the experiments are given in
Table 3. The gear pair has been subjected to 250 Nm opTable 3
Geometrical parameters of gears

Number of teeth
Pitch diameter (mm)
Tooth depth (mm)
Module
Pressure angle ( )
Distance of centres (mm)
Width (mm)

Test gear

Pinion

41
123
5.41
3
20
87
19

17
51

K. Aslantas, S. Tasgetiren / Wear 257 (2004) 11671175

1171

Fig. 5. The FZG test machine used in the pitting tests.

erational torque and the number of revolutions of pinion

was 1500 rev/min. Test gear tooth surfaces have been observed in per 5 105 revolution. When pits of the size about
0.5 mm have been observed on any tooth flank, the test run
was stopped. The pitting size values obtained experiments
are the mean of five gear teeth samples results. Pitting sizes
are measured approximately using microscope (50) and the
exact pitting sizes are measured using scanning electron microscope after the experiments.
3.3. Fatigue crack growth testing
Crack propagation testing is necessary to obtain the parameters of ParisErdogan equation. These are namely the
C and m values used in Eq. (9). For this purpose fatigue tests were conducted for the M(T) specimens with
a thickness of 4 mm, according to ASTM E647 standard.
The notch preparation is made by electrical-discharge machining. The crack propagation experiments are carried out
with a unaxial tensioncompression test machine (Instron
8501). All experiments are carried out at room temperature.
The loading frequency is selected as 15 Hz. The load ratio
(R) is 0.1. The specimens are clamped by hydraulic grips.
The crack length is observed using a travelling microscope
(30) during the crack propagation. The fatigue parameters
obtained from experimental analyses are given in Table 5.
Each value given in table is the mean value of at least three
samples.
Table 4
Gear pitting test results for maximum pitting size (mean value of three
samples)
Sample
A
B
As-cast

Average pitting formation life

(cycles 106 )
4.700
3.500
1.850

Pitting size (maximum)

(m)
400
460
500

4. Results and discussion

4.1. Experimental results
In the results of experimental studies, it is observed that
pitting failures occur roughly between the initial point of single tooth pair contact and pitch line. This was an expected
phenomenon. Because, this location of the tooth is subjected
to highest load due to the single tooth pair contact. The pitting
failure size observed on the tooth surface changes between
150 and 500 m at the pitch line. Maximum pitting size at
the tooth surface was taken into consideration to stop the experiment. Maximum pitting sizes on the tooth surface and the
average number of load cycle are given in Table 4.
Gear materials are generally subjected to heat treatment
to prevent surface fatigue failures such as pitting or spalling.
Spherical ductile iron alloyed with Ni and Cu has a good hardenability property when appropriate parameters are selected
for austempering process. Pitting formation life increases
when the austempering temperature decreases. Particularly,
the effect of austempering process on the pitting formation
life can be seen when comparing the gear austempered at
325 C with as-cast gear. Therefore, the highest hardness of
the 325 C was responsible for longest pitting formation life.
This feature can be the major advantage for gear applications.
4.2. Subsurface crack initiation and propagation
There are two stages, i.e., crack nucleation and propagation, in machine components subject to rolling contact loadTable 5
Fatigue parameters obtained from fatigue crack growth experiments

Sample
C (mm/cycle (MPa m))
A
B
As-cast

1012

1.06
3.79 1013
2.07 1014

m
5.1
6.5
7.01

1172

K. Aslantas, S. Tasgetiren / Wear 257 (2004) 11671175

4.3. Numerical results

Fig. 6. SEM micrograph of pitting failure on gear tooth surface (austempered

at 325 C for 60 min).

ing. Nodule is the crucial factor for crack initiation in austempered ductile iron materials. Dommarca et al. [16] stated that
the crack nucleation resistance of austempered ductile iron
is lower than that of steel. Because austempered ductile iron
materials behave as a metal matrix composite, cyclic loading causes to weaken the interface bond between the matrix
and the nodule. Therefore, microcracks emanate from the
graphite nodules and grow to link with other graphites.
The direction of microcrack changes depending on the
presence of nodules near the crack tip. For this reason, the
pitting failures have irregular form. The typically smooth
spall surface observed in bearing steel such as SAE 52100
[16] contrasts to the irregular one observed in austempered
ductile irons. Dommarca et al. have also observed that spall
failures occur as V-shape pointing opposite to rolling direction in SAE 52100. It can be seen from Figs. 5 and 6 that the
fracture surfaces of the pitting failures have irregular morphology in the materials studied in the present analysis. In
addition, Figs. 6 and 7 also show the subsurface crack propagation due to cyclic contact loadings in black zone indicated
with arrows.

Variation of the KI and KII at the crack tips are analysed

for different load positions. Friction is not considered because
the movement at the pitch line is only rolling. Rebbechi et
al. [17] measured the friction coefficient for this contact as
0.040.06 which may be regarded frictionless. Fifteen different positions from d = 0.35 to +0.35 mm are selected
for the load (see Fig. 3b). The mode I and II stress intensity
factors are computed by Eqs. (5)(7). Fig. 8 gives the results
of these analyses. KII have both negative and positive values
and this affects both the propagation direction and the propagation rate importantly according to maximum tangential
stress criterion. Presence of positive KII values deflects the
crack away from free surface. Negative KII values cause the
crack to deviate toward the surface. Maximum and minimum
values of KII occur at d = 0.15 and +0.15 mm load positions, respectively. Therefore, the subsurface crack deflects
from free surface at d = 0.15 mm and deviates towards the
free surface at d = +0.15 mm. Minimum KI value occurs at d
= 0 and values of KI are negative for all load positions.
4.4. Simulation of the crack propagation and pitting
formation
Pitting formation at the surface can be analysed using fracture/fatigue mechanics. In the majority of investigations in
fracture mechanics, it is assumed that the dependence between log (da/dN) and log K is linear and can be described
by the equation derived by Paris and Erdogan [18]:
da
= C[K]m
dN

where C and m are the material parameters that are determined experimentally. For a growing subsurface crack under
constant amplitude compressive loads, the conditions at the
crack tip are defined by the current value of K. For crack
growth due to mixed mode cyclic loading, K is defined by
Tanaka [19] as:
4
Keff
= KI4 + 8KII4

Fig. 7. SEM micrograph of pitting failure on gear tooth surface (as-cast gear
material).

(9)

(10)

where KI = KI,max KI,min and KII = KII,max KII,min .

KI is taken to be zero if both KI,max and KI,min have negative
values. KI = KI,max if KI,min < 0 and KI,max > 0 [20]. In the
majority of the studies about subsurface cracks, negative KI
is not considered because it has no effect on the crack growth
and KI,min is assumed zero [21,22].
In order to calculate the time elapsed for the crack to reach
to the surface, following procedures are practiced for each
loading cycle: Firstly KII values are determined for different
load positions, then the load positions occurring KII,max and
KII,min are determined for initial crack length. Secondly, the
crack length is increased step by step by a da increment and
new KII,max and KII,min values are determined by considering
the load position (Fig. 9).

K. Aslantas, S. Tasgetiren / Wear 257 (2004) 11671175

1173

Fig. 8. Variation of KI and KII at the left and right crack tips with respect to load position.

As seen from the Fig. 9, positive KII values decrease and

negative KII values increase when the crack length increases.
Negative KII values increase as four times relative to the initial
crack length when the right crack tip reaches the surface. In
addition, the load position which max and min KII values
occur do not depend on the crack length. As a result of these
considerations, Keff value in Eq. (10) is revised as:
4
= 8KII4
Keff

(11)

Keff values are calculated for each crack length increased

by da for every step and these are used in the ParisErdogan
equation. The increment da is used as 0.5 of initial crack
length. Therefore, Eq. (9) is rearranged as:
da
= C[Keff ]m
dN

(12)

The local finite element mesh around the initial crack is

shown in Fig. 10 (step 1). During finite element analysis, the
right crack tip is extended in the direction of the maximum

tangential stress. The incremental procedure is repeated until

the right crack tip reaches the contact surface. In Fig. 10, steps
411 illustrate the crack propagation towards the surface.
Step 15 depicts the crack propagation from the left crack tip
after the right tip has reached the surface. Numerical analyses
have shown that stress intensity factor at the right crack tip is
much higher than the left crack tip. Therefore, the right crack
tip will first start to propagate [10].
4.5. Comparison of the numerical and experimental
results
The predicted pitting size by the finite element analysis is
about 450 m. The pitting size observed on the tooth surfaces
after the FZG test changes between 400 and 500 m. Pitting
sizes and number of cycle for maximum pitting are given in
Table 6 for comparison. For the determination of the depth
of the failure, the teeth are cut to obtain the cross-section.
Fig. 11 shows SEM photographs of the pitting failure from
the cross-section. It is concluded that the maximum depth of
the failures reach up to 180 m. A finite element study carried
out by Aslantas et al. [23] has shown this value as 150 m.
This also gives crack nucleation depth of Hertz contact theory
[11]. The difference is caused due to existence of nodules.
The predicted subsurface crack propagation path and the test
results are shown in Fig. 11. As a result, the predicted crack
path and pitting size agree with experimental results in a very
good manner.
Table 6
Comparison of experimental and numerical results
Gear

Fig. 9. Variation of KII with respect to the crack length and the load position
in as-cast gear material for right crack tip.

Experiment

Numerical

Pitting size Number of cycles

(m)
(cycle 106 )

Pitting size
(m)

Number of cycles
(cycle 106 )

450

5.168
3.238
1.672

325 C 400
425 C 460
As-cast 500

4.700
3.500
1.850

1174

K. Aslantas, S. Tasgetiren / Wear 257 (2004) 11671175

Fig. 10. Fatigue crack growth of subsurface crack and variation of KII values at crack tips.

Fig. 11. SEM photograph of the cross-section, austempered at 325 C (a), as-cast (b) and numerically obtained pitting form of the tooth surface.

5. Conclusions
The present study considers the pitting failure model based
on the finite element analysis and linear elastic fracture mechanics. Also the effect of austempering process on the pitting formation in spur gear made of ductile iron is analysed.
A series of experimental study is carried out to determine the

pitting formation life on the tooth surface. Experimental and

numerical results from obtained analyses are compared.
In the experimental analysis, the pitting failure size observed on the tooth surface changes between average 150
and 500 m at the pitch line. Pitting formation life increases
when the austempering temperature decreases. So the pitting
resistance of the ductile iron gears are affected by austem-

K. Aslantas, S. Tasgetiren / Wear 257 (2004) 11671175

pering temperature and time. In addition, fracture surfaces of

the pitting failures have irregular morphology and the shapes
of pitting do not depend on rolling direction.
Numerical analysis used in the paper is appropriate for
determining the pitting resistance of the ductile iron gears.
Finite element results provide a good prediction tool both
for the formation life and pitting shape in a reasonable accuracy with the necessary material properties such as elasticity
modulus, Poissons ratio and the constant of ParisErdogan
equation which can be obtained by standard experimental
procedures if not available in the literature.
It can be concluded that, for the gears made from austempered ductile iron, the pitting failure time can be predicted
without testing of gears. When some standard experimental
values are known, the finite element method and fracture mechanics analyses can be utilized for the life prediction of the
gears.

References
[1] C.K. Lin, C.W. Chang, Influence of heat treatment on fatigue
crack growth of austempered ductile iron, J. Mater. Sci. 37 (2002)
709716.
[2] K. Aslantas, S. Tasgetiren, Edge spalling formation in a plate due
to moving compressive load, Turkish J. Eng. Env. Sci. 27 (2003)
333338.
[3] S. Tasgetiren, K. Aslantas, A numerical study of the behaviour of
surface cracks under dry-sliding conditions, Mater. Des. 24 (2003)
273279.
[4] C.H. Chue, H.H. Chung, Pitting formation under rolling contact,
Theor. Appl. Fract. Mech. 34 (2000) 19.
[5] L.M. Keer, M.D. Bryant, A pitting model for rolling contact fatigue,
J. Lubr. Technol. 105 (1983) 198205.
[6] S.D. Sheppard, D.A. Hills, J.R. Barber, An analysis of fretting
cracksII. Unloading and reloading phases, Int. J. Solids Struct.
22 (1986) 387396.
[7] D.A. Hills, M. Comninou, A normally loaded half-plane with an
edge crack, Int. J. Solids Struct. 21 (1985) 399410.

1175

[8] L. Magalhaes, J. Seabra, C. Sa, Experimental observations of contact

fatigue crack mechanisms for austempered ductile iron (ADI) discs,
Wear 246 (2000) 134148.
[9] L. Magalhaes, J. Seabra, Wear and scuffing of austempered ductile
iron gears, Wear 215 (1998) 237246.
[10] S. Glodez, J. Flasker, Z. Ren, A new model for the numerical determination of the pitting resistance of gear teeth flanks, Fatigue Fract.
Eng. Mater. Struct. 20 (1997) 7183.
[11] K.L. Johnson, Contact Mechanics, Cambridge University Press, London, 1985.
[12] C.L. Tan, Y.L. Gao, Treatment of bimaterial interface crack problems
using the boundary element method, Eng. Fract. Mech. 36 (6) (1990)
919932.
[13] FRANC2D Users Manual, available from: http://www.cfg.
cornell.edu.
[14] American Society for Testing and Materials, Standard test method
for evaluating the scuffing load capacity of oils (FZG visual method),
Annual Book of ASTM Standards 2001, vol. 06.04, USA, pp.
159164.
[15] Institute for Machine Elements Gear Research Centre (FZG):
http://www.fzg.mw.tum.de/index en.php4.
[16] R.C. Dommarca, P.C. Bastias, H.A. Dall, G.T. Hahn, C.A. Rubin,
Rolling contact fatigue (RCF) resistance of austempered ductile iron
(ADI), Wear 221 (1998) 6974.
[17] B. Rebbechi, F.B. Oswald, D.P. Townsend, Measurement of gear
tooth dynamic friction, NASA technical report ARL-TR-1165 (1996)
19.
[18] P.C. Paris, F. Erdogan, A critical analysis of crack propagation laws,
J. Basic Eng. 85 (1963) 528534.
[19] K. Tanaka, Fatigue crack propagation from a crack inclined to the
cyclic tensile axis, Eng. Fract. Mech. 6 (1974) 493507.
[20] K.S. Chan, Y.D. Lee, D.L. Davidson, S.J. Hudak, A fracture mechanics approach to high cycle fretting fatigue based on the worst
case fret concept, Int. J. Fract. 112 (2001) 299330.
[21] K. Komvopoulos, Subsurface crack mechanisms under indentation
loading, Wear 199 (1996) 923.
[22] K. Komvopoulos, S.S. Cho, Finite element analysis of subsurface
crack propagation in a half-space due to moving asperity contact,
Wear 209 (1997) 5768.
[23] K. Aslantas, S. Tasgetiren, K. Dundar, Investigation of the effect of
friction force on damage formation in spur gears, Pamukkale Univ.
J. Eng. Sci. 8 (1) (2002) 3340.